Download Emory University School of Law

Emory University School of Law
Public Law & Legal Theory Research Paper Series
Research Paper No. 05-25
FORM AND (DYS)FUNCTION IN SEXUAL HARASSMENT LAW:
BIOLOGY, CULTURE, AND THE SPANDRELS OF TITLE VII
Julie A. Seaman
Emory University School of Law
This paper can be downloaded without charge from:
The Social Science Research Network Electronic Paper Collection:
http://papers.ssrn.com/abstract=758509
FORM AND (DYS)FUNCTION IN SEXUAL
HARASSMENT LAW: Biology, Culture, and the
Spandrels of Title VII
Julie A. Seaman†
ABSTRACT: The question of sex difference has long divided feminists,
social scientists, policymakers, and legal academics. Most recently, the
issue has resurfaced as prominent legal scholars have relied on
evolutionary arguments to suggest that men and women might be different
in ways that are relevant to sex discrimination law in general and to sexual
harassment law in particular. At the same time, the question of causation
under Title VII, which requires that discrimination be “because of” a
plaintiff’s sex in order to be actionable, has assumed central importance in
sexual harassment doctrine. This article proposes that the very evolutionary
theories advanced by critics of Title VII sex discrimination doctrine in fact
support the view that the typical behavior patterns seen in sexual
harassment cases occur “because of” the plaintiff’s sex. Furthermore, this
article argues that, under current Supreme Court jurisprudence setting out
the contours of employer liability for harassing behavior by employees,
employers should be liable for harassment that results where the employer
allows, fosters, or fails to correct workplace conditions that are likely to
give rise to the typical harassment behaviors analyzed herein. Though
often perceived as being in conflict, biological explanations of many human
behaviors and feminist conceptions of the causes and harms of workplace
sexual harassment share much common ground. This article seeks to
explore that common ground and, in so doing, to offer constructive
solutions to specific doctrinal and societal problems.
†
Assistant Professor, Emory University School of Law. J.D. Harvard Law School. I am
grateful to many people who offered their time and their thoughts in connection with previous
drafts of this article. In particular, I wish to thank Howard Abrams, Martha Fineman, Rich
Freer, Bill Mayton, Patricia Adair Gowaty, Marc Miller, Michael Perry, Jennifer Romig, Paul
Rubin, Ani Satz, Charlie Shanor, Sara Stadler, and participants at the Emory Law School
faculty workshop at which I presented an earlier version of this article. I also wish to thank
participants at the 7th Annual SEAL Conference for most thoughtful and helpful questions and
comments. Owen Jones was especially generous with his time, expertise, and wisdom. Finally,
Kingsley Browne provided insightful and incisive comments and made me thankful to be part of
an intellectual community in which disagreement coexists with discussion. It should go without
saying that my mention of their names in no way implies that any of these individuals endorses
the views set out in this article. Finally, I am grateful for the wonderful research assistance
provided by Sarah Kemmerer, Bethany Kohl, and Jordan Reifler.
322
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
I. INTRODUCTION......................................................................................323
II. APPLICATION TO SELECTED CONTEXTS ................................................334
A. Preexisting Sexualized Atmosphere in the Workplace ...................334
B. Same-Sex Male on Male Harassment ............................................335
C. One or a Few Women in a Predominantly Male Workplace .........335
III. THE DEVELOPMENT WARS: BEYOND GENES VERSUS
ENVIRONMENT ......................................................................................336
A. Arguments from Nature: Natural Selection and Sexual
Selection .........................................................................................340
1. Sex Difference Through an Evolutionary Lens .......................347
2. Female Choice in Biology and Evolution
—Emerging Theories ...............................................................353
B. Arguments from Nurture: Theories of Social Construction...........355
C. Nature and Nurture in the Law of Sexual Harassment ..................361
1. Sex Differences in Cognitive Ability and Preferences ............368
2. Evolutionary Arguments and Hostile Work Environment
Sexual Harassment Law...........................................................375
3. Summary ..................................................................................378
IV. EVOLUTIONARY EXPLANATIONS FOR TYPICAL HARASSMENT
BEHAVIORS ...........................................................................................384
A. Recurring Patterns in Harassment Cases ......................................385
B. Evolutionary Explanations.............................................................388
1. “Give Me What I Want” ..........................................................389
2. “Get Out of My Space” ............................................................397
3. “Take It Like a Man” ...............................................................402
4. The Significance of the Workplace Context Under
Evolutionary Theory ................................................................405
V. DOCTRINAL SPANDRELS IN THE LAW OF SEXUAL HARASSMENT ..........410
A. The Terminology Spandrel: What is “Sex?” .................................412
B. The “Mother’s Eyes, Father’s Nose” Spandrel: Who Are the
Parents of Hostile Environment Sexual Harassment Law? ...........416
1. The Analogy to Race................................................................417
2. The Feminist Anti-Subordination Argument ...........................421
C. The Causation Spandrel: Whose Intention Counts? ......................425
VI. CONCLUSION .........................................................................................432
37:0321]
SEXUAL HARASSMENT LAW
323
The great central dome of St Mark’s Cathedral in Venice
presents in its mosaic design a detailed iconography expressing
the mainstays of Christian faith . . . . Each quadrant meets one of
the four spandrels in the arches below the dome. Spandrels—the
tapering triangular spaces formed by the intersection of two
rounded arches at right angles—are necessary architectural byproducts of mounting a dome on rounded arches . . . . The design
is so elaborate, harmonious and purposeful that we are tempted to
view it as the starting point of any analysis, as the cause in some
sense of the surrounding architecture. But this would invert the
proper path of analysis.1
I. INTRODUCTION
The conflict between those who would explain human behavior as a
product of genes and those who would explain it as a product of
environment is an old one. Recently, however, the debate has garnered
renewed public interest with the “mapping” of the human genome2 and
stories in the press about genes “for” traits and behaviors from shyness to
depression to homosexuality.3 The debate is reflected, too, in the legal
1.
Stephen Jay Gould & Richard C. Lewontin, The Spandrels of San Marco and the
Panglossian Paradigm: A Critique of the Adaptationist Programme, 205 PROC. OF THE ROYAL
SOC. OF LONDON 581, 581–82 (1979).
2.
See MATT RIDLEY, GENOME: THE AUTOBIOGRAPHY OF A SPECIES IN 23 CHAPTERS, at 2
(2000) (“On June 26, 2000, President Bill Clinton in the White House and Tony Blair in
Downing Street simultaneously announced that the rough draft [of the entire human genome]
was complete.”). For examples of contemporaneous media accounts, see Sharon Begley, Just
How Many Genes Does It Take to Make a Human? Wanna Bet?, WALL ST. J., May 23, 2003, at
B1; Claire Fraser, Who’s Next, NEWSWEEK, Dec. 31, 2001, at 83; Bernadette Tansey, Genome
Project Completes Map of Human DNA, S.F. CHRON., Apr. 15, 2003, at A2; Nicholas Wade,
Gene-Mappers Take New Aim at Diseases, N.Y. TIMES, Oct. 30, 2002, at A23; Unraveling the
Genome, NEWSWEEK, June 24, 2002, at 76.
3.
E.g., Mike Bygrave, The Year of the Gene: False Dawns in the Brave World of New
Genetics, OBSERVER, Dec. 22, 2002, at 20; Benedict Carey, Payback Time: Why Revenge Tastes
So Sweet, N.Y. TIMES, July 27, 2004, at F1 (“[T]he urge to extract a pound of flesh, researchers
find, is primed in the genes.”); Gene Linked to Shyness, TIMES (London), Apr. 15, 2003, at
Public Agenda 6; Mental Health: Genetic Differences Partly Account for Higher Incidence of
Depression in Women, GENOMICS & GENETICS WKLY., Aug. 8, 2003, at 34 (discussing findings
regarding a gene “for” depression in women); “Shyness Gene” Discovered, HEALTH NEWSWIRE
CONSUMER, Apr. 8, 2003; Environmental Factors Triggering Illness and What Scientists are
Doing to Identify the Genes That Contribute to Depression (NPR radio broadcast, July 18,
2003) (discussing a “gene for” depression). The 2003 Pulitzer Prize for literature was recently
awarded to a novel depicting a genetically male person with a rare recessive genetic mutation
that results in his being born with ambiguous outward genitalia and raised as a female and then,
324
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
academy and in the contours of the law itself. The recent law and evolution
scholarship, on the one hand, and some feminist legal scholarship, on the
other, tend to approach legal issues from the contrasting perspectives of
either biology or culture.4 In addition, courts routinely rely on some
conception of “human nature” in formulating and applying legal rules.5
at puberty, experiencing a crisis of identity. See generally JEFFREY EUGENIDES, MIDDLESEX
(2002). The story has close factual parallels with an actual case of a genetic male raised as a
female. See JOHN COLAPINTO, AS NATURE MADE HIM: THE BOY WHO WAS RAISED AS A GIRL
(2000) (describing the identity crisis and its effects); see also MATT RIDLEY, THE RED QUEEN:
SEX AND THE EVOLUTION OF HUMAN NATURE 258 (2003) (describing 5-alpha-reductase
deficiency, the rare genetic condition that underlies the events fictionalized in Jeffrey
Eugenides’s novel). Sadly, the person described in Colapinto’s book recently committed
suicide. John Colapinto, Gender Gap: What Were the Real Reasons Behind David Reimer’s
Suicide?, at http://slate.msn.com/id/2101678/ (last visited Feb. 20, 2005).
4.
For examples of law and evolution scholarship, see generally JOHN H. BECKSTROM,
EVOLUTIONARY JURISPRUDENCE: PROSPECTS AND LIMITATIONS ON THE USE OF MODERN
DARWINISM THROUGHOUT THE LEGAL PROCESS (1989); KINGSLEY R. BROWNE, BIOLOGY AT
WORK: RETHINKING SEXUAL EQUALITY (2002) [hereinafter BROWNE, BIOLOGY AT WORK];
RICHARD A. POSNER, SEX AND REASON (1992); Kingsley R. Browne, An Evolutionary
Perspective on Sexual Harassment: Seeking Roots in Biology Rather Than Ideology, 8 J.
CONTEMP. LEGAL ISSUES 5 (1997) [hereinafter Browne, Seeking Roots]; Kingsley R. Browne,
Biology, Equality, and the Law: The Legal Significance of Biological Sex Differences, 38 SW. L.
J. 617 (1984) [hereinafter Browne, Biology, Equality]; Kingsley R. Browne, Sex and
Temperament in Modern Society: A Darwinian View of the Glass Ceiling and the Gender Gap,
37 ARIZ. L. REV. 971 (1995) [hereinafter Browne, Sex and Temperament]; Kingsley R. Browne,
Women at War: An Evolutionary Perspective, 49 BUFF. L. REV. 51 (2001) [hereinafter Browne,
Women at War]; Charles C. Crawford & Marc A. Johnston, An Evolutionary Model of
Courtship and Mating as Social Exchange: Implications for Rape Law Reform, 39 JURIMETRICS
J. 181 (1999); Richard A. Epstein, Two Challenges for Feminist Thought, 18 HARV. J.L. & PUB.
POL’Y 331 (1995) [hereinafter Epstein, Two Challenges]; Gertrud M. Fremling & Richard A.
Posner, Status Signaling and the Law, With Particular Application to Sexual Harassment, 147
U. PA. L. REV. 1069 (1999); Owen D. Jones, Evolutionary Analysis in Law: An Introduction
and Application to Child Abuse, 75 N.C. L. REV. 1117 (1997) [hereinafter Jones, Child Abuse];
Owen D. Jones, Law and Biology: Toward an Integrated Model of Human Behavior, 8 J.
CONTEMP. LEGAL ISSUES 167 (1997) [hereinafter Jones, Law and Biology]; Owen D. Jones,
Law, Emotions, and Behavioral Biology, 39 JURIMETRICS J. 283 (1999); Owen D. Jones,
Reproductive Autonomy and Evolutionary Biology: A Regulatory Framework for TraitSelection Technologies, 19 AM. J.L. & MED. 187 (1993); Owen D. Jones, Sex, Culture, and the
Biology of Rape: Toward Explanation and Prevention, 87 CAL. L. REV. 827 (1999) [hereinafter
Jones, Rape]; John Monahan, Could “Law and Evolution” Be the Next “Law and Economics”?,
8 VA. J. SOC. POL’Y & L. 123 (2000); Jeffrey Evans Stake, Can Evolutionary Science
Contribute to Discussions of Law?, 41 JURIMETRICS J. 379 (2001). See also Richard A. Epstein,
Liberty, Patriarchy, and Feminism, 1999 U. CHI. LEGAL F. 89, 105–112 [hereinafter Epstein,
Liberty, Patriarchy] (discussing the issue of differential outcomes and suggesting that they can
be largely explained by differing average endowments and psychologies on men and women
that are rooted in biological differences). For examples of feminist legal scholarship focusing on
culture and minimizing biology, see CATHARINE A. MACKINNON, TOWARD A FEMINIST THEORY
OF THE STATE 224 (1989) (“Clearly, there are many differences between women and men.
Systematically elevating one-half of a population and denigrating the other half would not likely
37:0321]
SEXUAL HARASSMENT LAW
325
produce a population in which everyone is the same.”); see also Kathryn Abrams, Social
Construction, Roving Biologism, and Reasonable Women: A Response to Professor Epstein, 41
DEPAUL L. REV. 1021, 1023 (1992) (reviewing social constructionist feminist theories of
difference, and arguing that these theories offer “a useful corrective to many popular, and some
scholarly, conceptions” of biological or “natural” sex difference); Mary Becker, Patriarchy and
Inequality: Towards a Substantive Feminism, 1999 U. CHI. LEGAL F. 21, 23 (“Social structures
and the individuals within them create and reproduce inequalities linked to sex, race, class,
religion, ethnicity, and other ‘differences.’”); Vicki Schultz, Reconceptualizing Sexual
Harassment, 107 YALE L.J. 1683, 1686 n.8 (1998):
I use the terms “gender” and “sex” interchangeably . . . to refer to the
complex process of socializing human beings into the identities of men and
women, to the element of social relationships based on differences that
society attributes to people with those two identities, and to the process of
signifying power through those identities.
Id.; Joan C. Williams, Deconstructing Gender, 87 MICH. L. REV. 797, 826 (1989) (“In fact, both
discrimination against women and women’s ‘choices’ must be seen as elements of an integrated
system of power relations that systematically disadvantages women.”).
5.
A recent Westlaw search for the term “human nature” in the state and federal court
databases resulted in more than 3,000 cases since 1998. See, e.g., Manson v. Brathwaite, 432
U.S. 98, 134 (1977) (Marshall, J., dissenting) (“[I]t is deeply ingrained in human nature to agree
with the expressed opinions of others—particularly others who should be more
knowledgeable—when making a difficult decision.”); Andersen v. United States, 170 U.S. 481,
510 (1898) (“The law in recognition of the frailty of human nature, regards a homicide
committed under the influence of sudden passion, or in hot blood, produced by adequate cause,
and before a reasonable time has elapsed for the blood to cool, as an offense of a less heinous
character than murder.”); Alvarez-Machain v. United States, 331 F.3d 604, 664 (9th Cir. 2003)
(Gould, J., dissenting) (“Human nature being what it is, and judicial nature following human
nature, it is only natural that well-meaning judges will desire to comment on important affairs of
the day involving political relations with other nations.”); Newport News Shipbuilding & Dry
Dock Co. v. Pounders, 326 F.3d 455, 459 (4th Cir. 2003) (“Human nature as it is, company
doctors, however honest, are likely to give the close calls to those who pay their salaries.”);
Lansdale v. Hi-Health Supermart Corp., 314 F.3d 355, 359 (9th Cir. 2002) (“It is not at all
surprising that [the plaintiff] wants even more than Congress provided; that is just the working
out of one of human nature’s quotidian drives. However, she must be content with her six figure
judgment, faute de mieux.”); Cal. First Amendment Coalition v. Woodford, 299 F.3d 868, 877
(9th Cir. 2002):
Finally, public observation of executions fosters the same sense of
catharsis that public observation of criminal trials fosters. Although this may
reflect the dark side of human nature, the Supreme Court has recognized that
the public must be permitted to see justice done, lest it vent its frustration in
extralegal ways.
Id.; Raheem v. Kelly, 257 F.3d 122, 138 (2d Cir. 2001) (“Further, it is human nature for a
person toward whom a gun is being pointed to focus his attention more on the gun than on the
face of the person pointing it.”); CH2M Hill, Inc. v. Herman, 192 F.3d 711, 713 (7th Cir. 1999)
(“Whenever accidental death occurs, it is human nature to place blame.”); Paolella v. BrowningFerris, Inc., 973 F. Supp. 508, 514 (E.D. Pa. 1997) (“It is a matter of common sense, and of
general knowledge of human nature, that people are not inclined knowingly to consent to being
economically gouged.”); United States v. Gorayska, 482 F. Supp. 576, 582 n.10 (S.D.N.Y.
1979) (“The choice of false confession is voluntary, but the false confession is associated with a
prospect (namely, of escape from present harm) so tempting that it is not in human nature to
resist it.”). Jurisprudes, of necessity, confront the issue as the foundation of their thinking:
326
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
Workplace sexual harassment presents an ideal arena in which to
examine the interaction among law, biology, and cultural understanding. It
is a phenomenon that generally involves behavior between males and
females.6 It involves patterns of behavior implicating sex, aggression, and
resource acquisition that are of primary interest to behavioral biologists. It
involves patterns of subordination and dominance, both sexual and
economic, of primary interest to critical scholars who focus on entrenched
cultural patterns and sex roles. It raises questions of sex and gender
difference that are of central concern to evolutionary biologists, feminists,
legal scholars across various areas, and humans in general. It is notoriously
resistant to easy answers, either legal or social. In sum, sexual harassment
presents a nearly perfect crucible in which simmer the incendiary questions
of sex, gender, biology, culture, and law.
The Supreme Court’s pronouncement in 1986 that sexual harassment
was a form of sex discrimination7 prohibited by Title VII of the Civil Rights
“What is a human being? Legal theorists must, perforce, answer this question: jurisprudence,
after all, is about human beings.” Robin West, Jurisprudence and Gender, 55 U. CHI. L. REV. 1,
1 (1988).
6.
The vast majority of sexual harassment complaints concern male harassment of
females. See U.S. EQUAL EMPLOYMENT OPPORTUNITY COMM’N, TRENDS IN HARASSMENT
CHARGES FILED WITH THE EEOC DURING THE 1980S AND 1990S, http://www.eeoc.gov/stats/
harassment.html (last visited Jan. 6, 2005) (finding that, between 1992 and 1998, approximately
one-tenth of harassment complaints were filed by men); Ann Juliano & Stewart J. Schwab, The
Sweep of Sexual Harassment Cases, 86 CORNELL L. REV. 548, 560 (2001) (finding, in their
comprehensive survey of all published sexual harassment decisions in federal district courts
between 1986 and 1995, that men comprised only 5.4% of plaintiffs). The more exceptional or
marginal cases have, however, been the basis for much of the scholarly debate which has been
instrumental in providing the theoretical underpinnings of the doctrine. See Mary Anne C. Case,
Disaggregating Gender from Sex and Sexual Orientation: The Effeminate Man in the Law and
Feminist Jurisprudence, 105 YALE L.J. 1, 3 (1995) (“[Q]uite apart from the concerns we have
for men, particularly effeminate men, in and of themselves, it is important for women and
feminists to concern themselves with the treatment of the effeminate man.”); Katherine M.
Franke, What’s Wrong with Sexual Harassment?, 49 STAN. L. REV. 691, 695 (1997) (“The
disjunctive doctrine that has emerged from the courts’ difficulty in dealing with the same-sex
cases provides a timely opportunity and excuse for reexamining, reaffirming, and updating
feminist conceptions of sexual harassment as a form of sexual discrimination.”); cf. Schultz,
supra note 4, at 1774–89 (demonstrating that her proposed “competence-centered” paradigm
provides a principled way of deciding the same-sex cases, based upon whether the male
harassers have engaged in “harassment that drives away men who fail to conform to the desired
image of masculinity or that incorporates them as weak and inferior workers”). Whether one
views the same-sex or female-harassing-male cases as relatively anomalous, or rather as central
to the theoretical project of building a foundation for the more usual cases, depends in large part
upon one’s conception of the causes and harms of sexual harassment.
7.
See Meritor Sav. Bank v. Vinson, 477 U.S. 57, 63 (1986). The Court held, in a part of
its opinion with which all nine justices concurred, that “when a supervisor sexually harasses a
subordinate because of the subordinate’s sex, that supervisor ‘discriminate[s]’ on the basis of
37:0321]
SEXUAL HARASSMENT LAW
327
Act of 19648 raised more questions than it answered. In subsequent years,
the Court has moved, in fits and starts, toward a coherent explanation of
when an employer is liable under Title VII for abusive working conditions
created by coworkers and/or supervisors of a complaining employee.9 In a
unanimous decision addressing the issue of same-sex harassment, the Court
stressed that, in all sexual harassment cases, proof that the harassing
behavior was based on the plaintiff’s sex is critical.10 At the same time,
feminist legal scholars have attempted to make sense of the wrong of sexual
harassment and to explain precisely why it constitutes discrimination
“because of . . . sex” as required under Title VII.11 And social scientists and
sex.” Id. at 65 (alteration in original). Indeed, the Court noted that even the petitioner in the case
did not challenge this general proposition. Id.
8.
42 U.S.C. § 2000e-2 (2000).
9.
The Supreme Court has considered major substantive issues concerning sexual
harassment sex discrimination under Title VII in six decisions, three of which were decided
without dissent. See Pa. State Police v. Suders, 542 U.S. 129, 134 (2004) (holding that
constructive discharge effectuated by hostile environment sexual harassment precludes
employer from asserting affirmative defense only where “the plaintiff quits in reasonable
response to an employer-sanctioned adverse action officially changing her employment status or
situation”); Burlington Indus., Inc. v. Ellerth, 524 U.S. 742, 754 (1998) (clarifying standards of
employer liability for discriminatory actions of supervisors); Faragher v. City of Boca Raton,
524 U.S. 775, 806 (1998) (same); Oncale v. Sundowner Offshore Servs., Inc., 523 U.S. 75, 78–
80 (1998) (holding that same-sex sexual harassment is actionable under Title VII); Harris v.
Forklift Sys., Inc., 510 U.S. 17, 21–22 (1993) (holding that plaintiff need not demonstrate
serious psychological harm in order to state a claim for hostile environment discrimination, and
clarifying standards under which offensive behavior is actionable under Title VII); Meritor, 477
U.S. at 66 (recognizing that hostile environment harassment may give rise to liability for sex
discrimination under Title VII).
10. See Oncale, 523 U.S. at 81 (“Whatever evidentiary route the plaintiff chooses to
follow, he or she must always prove that the conduct at issue was not merely tinged with
offensive sexual connotations, but actually constituted ‘discrimina[tion] . . . because of . . .
sex.’”) (alterations in original); see also id. at 82 (Thomas, J., concurring) (“I concur because
the Court stresses that in every sexual harassment case, the plaintiff must plead and ultimately
prove Title VII’s statutory requirement that there be discrimination ‘because of . . . sex.’”).
11. Under § 703 of Title VII, it is “an unlawful employment practice for an employer . . .
to discriminate against any individual with respect to his compensation, terms, condition, or
privileges of employment, because of such individual’s race, color, religion, sex, or national
origin.” 42 U.S.C. § 2000e-2(a)(1) (2004). For a critical discussion of the various theories upon
which causation has been based, and the articulation of an alternative theory, see generally
Franke, supra note 6. See also Case, supra note 6, at 3 (arguing that a wide spectrum of Title
VII cases can, and should, be understood as caused by gender discrimination; that is, that
employers impose upon both men and women requirements to conform to stereotypically
masculine or feminine roles and that such requirements place women in traditionally masculine
jobs in the “double bind” of simultaneous expectations of both masculine and feminine
behavior); Robert A. Kearney, The Unintended Hostile Environment: Mapping the Limits of
Sexual Harassment Law, 25 BERKELEY J. EMP. & LAB. L. 87, 99–104 (2004) (criticizing the
approaches of Professors Franke and Schwartz for eliminating the statutorily required element
of intent in disparate treatment cases); David S. Schwartz, When is Sex Because of Sex? The
328
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
evolutionary psychologists have sought to measure and explain the
behaviors and psychologies that underlie the phenomenon of sexual
harassment in the workplace.12
Recently, some Title VII litigants and legal scholars have chipped away
at the veneer of the assumption of male-female equality to assert overtly
what had theretofore lurked below the surface: that men and women are
different in ways that are relevant to antidiscrimination law in general and
to workplace sexual harassment law in particular. They have grounded this
assertion on theories of sex difference, including but not limited to
biological theories of difference. Though few courts have relied explicitly
on such theories to find in favor of defendants,13 it is likely that notions
about male and female roles, temperament, and cognitive strengths and
weaknesses influence decisions in more subtle ways. In addition, it is likely
that these assumptions influence the decisions and behavior of employers
and employees in the workplace. Because laws aimed at altering entrenched
behavior patterns14 are bound to fail absent a concurrent change in social
Causation Problem in Sexual Harassment Law, 150 U. PA. L. REV. 1697, 1700 (2002)
(proposing that sexual words and conduct should be held to be “because of . . . sex” as a matter
of law); cf. Kenji Yoshino, Covering, 111 YALE L.J. 769, 772 (2002) (describing similar double
bind in the context of gay cultural assimilation and the pressure on gays to “cover”—to “make[]
it easy for others to disattend [their sexual] orientation.”)
12. See generally ROSABETH MOSS KANTER, MEN AND WOMEN OF THE CORPORATION
(1977); Sarah E. Burns, Issues in Workplace Sexual Harassment Law and Related Social
Science Research, 51 J. SOC. ISSUES 193 (1995); Barbara A. Gutek, Understanding Sexual
Harassment at Work, 6 NOTRE DAME J.L. ETHICS & PUB. POL’Y 335 (1992); Barbara A. Gutek
& Bruch Morasch, Sex-Role Spillover, and Sexual Harassment of Women at Work, 38 J. SOC.
ISSUES 55 (1982); Margaret S. Stockdale et al., Acknowledging Sexual Harassment: A Test of
Alternative Models, 17 J. BASIC & APPLIED SOC. PSYCHOL. 469 (1995); David E. Terpstra &
Douglas D. Baker, A Hierarchy of Sexual Harassment, 121 J. PSYCHOL. 599 (2002).
13. There have been some notable (and widely discussed) exceptions. See Equal
Employment Opportunity Comm’n (EEOC) v. Sears, Roebuck & Co., 628 F. Supp. 1264, 1305
(N.D. Ill. 1986), aff’d., 839 F.2d 302 (7th Cir. 1988). See generally Vicki Schultz & Stephen
Petterson, Race, Gender, Work, and Choice: An Empirical Study of the Lack of Interest Defense
in Title VII Cases Challenging Job Segregation, 59 U. CHI. L. REV. 1073 (1992) (analyzing
judicial treatment of the “lack of interest defense,” in both sex and race discrimination cases,
where the alleged lack of interest stems from either cultural, social, or biological forces); Vicki
Schultz, Telling Stories About Women and Work: Judicial Interpretations of Sex Segregation in
the Workplace in Title VII Cases Raising the Lack of Interest Argument, 103 HARV. L. REV.
1749 (1990) (analyzing Supreme Court decisions addressing lack of interest doctrine and how
courts have drawn the boundary between “coercion” and “choice”); Williams, supra note 4, at
813–21 (discussing EEOC v. Sears and the perils of modern domesticity).
14. See Michael L. Selmi, Sex Discrimination in the Nineties, Seventies Style: Case
Studies in the Preservation of Male Norms, working paper abstract, at
http://ssrn.com/abstracts=453485 (last revised Feb. 19, 2005) (showing that patterns of
egregious and blatant employment discrimination are still the norm, and not more subtle,
unconscious discrimination as many had hypothesized); see also Owen D. Jones, Law and the
37:0321]
SEXUAL HARASSMENT LAW
329
norms and attitudes,15 it is important to address theories about sex
differences rather than ignore them.
This article focuses on the issue of hostile environment sexual
harassment, and proposes that the insights of evolutionary science,
combined with those of feminist legal theory, can help to explain why the
patterns of behavior typically seen in workplace hostile environment cases
amount to discrimination “because of” sex in violation of Title VII. An
evolutionary perspective suggests that many of the typical patterns of
behavior aimed at soliciting sex, at status competition and exclusion, and at
exclusion of women from predominantly or traditionally male workplaces,
are ultimately based on the target’s sex. Furthermore, much feminist legal
analysis of sexual harassment is fully consistent with these explanations of
the underlying causes of sexual harassment. Finally, recent work in
evolutionary biology that focuses on the role of female choice in the
coevolution of male and female characteristics, and the implications of
constraints on female choice to the model of sexual selection, strongly
implies that the workplace context is highly relevant to understanding both
the causes of, and the harms that result from, sexual harassment.
Furthermore, an evolutionarily-informed explanation of the gendered
nature of workplace harassment16 fits quite easily into the existing doctrinal
framework of Title VII, once the fundamental structure of that framework is
understood. To reveal the primary, weight-bearing aspects of that structure
and to separate these from the theoretically more peripheral doctrinal
elements, I borrow from the debate in evolutionary biology the concept of
Biology of Rape: Reflections on Transitions, 11 HASTINGS WOMEN’S L.J. 151, 151–54 (2000)
(noting that despite necessary and well-intentioned legal reforms aimed at increasing deterrence,
reporting, and punishment, “rape reforms have had far less impact than hoped”); Jones, Rape,
supra note 4, at 830 & n.5 (same) (citing research generally reaching this conclusion).
15. Cf. Book Note, Real Reform?, 101 HARV. L. REV. 1978, 1979–83 (1988) (reviewing
SUSAN ESTRICH, REAL RAPE (1987)) (arguing that Estrich, by presenting the “easy” cases in an
emotionally persuasive way, fails to address the entrenched underlying social norms that would
have to change in order for her proposed legal reform to be effective). For an argument that law
is largely reactive to social change, see MARTHA ALBERTSON FINEMAN, THE NEUTERED
MOTHER, THE SEXUAL FAMILY, AND OTHER TWENTIETH CENTURY TRAGEDIES 11, 15–16 (1995)
(“Law is a conservative discipline, not easily used for ends that undermine the status quo.”); see
also LARRY D. BARNETT, LEGAL CONSTRUCT, SOCIAL CONCEPT: A MACROSOCIOLOGICAL
PERSPECTIVE ON LAW 13 (1993); AUGUSTUS B. COCHRAN III, SEXUAL HARASSMENT AND THE
LAW: THE MECHELLE VINSON CASE 191–208 (2004); Lawrence M. Friedman, Law, Lawyers,
and Popular Culture, 98 YALE L.J. 1579, 1583–84 (1989); Madhavi Sunder, Cultural Dissent,
54 STAN. L. REV. 495, 507 (2001).
16. Cf. FINEMAN, supra note 15, at 12, 49–54 (describing the concept of a “gendered life,”
which “references the actual or potential situations and circumstances women as a group may
encounter in our society,” and which “distinguish[es] women’s ‘reality’ from men’s in our
society”).
330
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
evolutionary “spandrels”17 and apply that concept to the law of sexual
harassment. As an evolutionary and biological metaphor, the architectural
term “spandrel” has proven illuminating, if controversial.18 In architecture,
the term refers to a space necessarily created by the architect’s functional
design decisions. Thus, the decision to mount a dome on four arches entails,
as a byproduct, four tapering triangular spaces between the sides of the
arches and the bottom of the dome.19 In 1979, Harvard scientists Stephen
Jay Gould and Richard Lewontin published a now-famous paper in which
they used the architectural analogy to criticize what in their view was the
tendency of evolutionary biologists to explain every atomistic trait of an
organism as an evolutionary adaptation.20
Regardless of the correct view of evolutionary adaptation, the spandrel
as a metaphor for structural byproducts presents intriguing possibilities for
the examination of complex legal doctrine. In the arena of Title VII sexual
harassment law, the spandrel metaphor is especially apt because of the
complicated way that current doctrine has evolved. The law of sexual
harassment was born of two very disparate jurisprudential parents: an
analogy to hostile racial environment discrimination on the one hand, and
17. In architecture, a spandrel is “an area between the extradoses of two adjoining arches,
or between the extrados of an arch and a perpendicular through the extrados at the springing
line.” RANDOM HOUSE WEBSTER’S UNABRIDGED DICTIONARY 1828 (2d ed. 1998). Spandrel as a
metaphor thus refers to a feature that appears to be designed or functional but actually emerged
only as the byproduct of other functional or designed features. The spandrels of the Cathedral of
San Marco in Venice might seem to have been a functional part of the design of the cathedral,
but in fact they are only the byproduct of the architecturally necessary arches. They are the
triangular spaces, albeit lovely, between the arches. See Gould & Lewontin, supra note 1.
According to this analysis, we might admire these triangular features of the cathedral, but we
should not pretend that they were purposely put there to serve a particular function in the
building. See id. In the same way, there are spandrels that have resulted from the evolution of
the law, but which did not themselves evolve to serve some independent function.
18. See, e.g., DANIEL C. DENNETT, DARWIN’S DANGEROUS IDEA 267–68 (1995) (“[Gould
and Lewontin] introduced another term, ‘spandrel,’ which has proven to be a highly successful
coinage in one sense: it has spread through evolutionary biology and beyond.”); Stephen Jay
Gould, The Exaptive Excellence of Spandrels as a Term and Prototype, 94 PROC. NAT’L ACAD.
SCI. U.S. 10750, 10750 (1997) [hereinafter Gould, Term and Prototype] (responding to
criticisms); Stephen Jay Gould, Fulfilling the Spandrels of World and Mind, in UNDERSTANDING
SCIENTIFIC PROSE 310, 325 (Jack Selzer ed., 1993).
19. When the arches are perpendicular to one another, as they are in the Cathedral of San
Marco in Venice which Gould and Lewontin use as the primary illustration for their analogy,
the spaces that are left between the arches exist in three dimensions rather than two.
Technically, the triangular wall spaces thus formed are called “pendentives.” There is some
disagreement as to whether these pendentives are a subset of the category “spandrels.” As
Gould points out in a later article revisiting his spandrel argument, critiques based on this
terminological rigidity are, in terms of evolutionary argument, utterly beside the point. See
Gould, Term and Prototype, supra note 18.
20. See Gould & Lewontin, supra note 1.
37:0321]
SEXUAL HARASSMENT LAW
331
the radical feminist critique of sexuality-as-dominance on the other.21 It is
possible, indeed likely, that the functional necessities of the doctrinal or
theoretical structures in each of these areas might have produced doctrinal
byproducts that are nonfunctional, or perhaps even dysfunctional, in the
current environment. In this article, I evaluate Title VII sexual harassment
doctrine from this perspective, and ask whether certain elements are in fact
“legal spandrels.”22 Because the inquiry is metaphorical rather than tangible,
it leaves much room for a decision to discard those doctrinal appendages
that are found to be dysfunctional in the current environment.23
In order to situate the evolutionary and legal analysis within a factual
context, the Article begins in Part II with a series of snapshots of potentially
actionable harassment scenarios. Following each hypothetical case is a brief
summary of the way in which the approach proposed in this Article would
apply, and also the way in which this approach might differ from the current
doctrinal and theoretical framework.
Part III begins with a description of the terms of the debate over the
respective influences of biology and culture upon human behavior in
general, and upon male-female differences in behavior in particular. It first
lays out the evolutionary, biological approach to sex difference. That
approach is then compared to the social construction explanation of
differences between men and women, after which focus shifts to the
scholarly debate in the legal literature about the existence, meaning, cause,
and relevance of average differences between men and women in
preference, ability, and cognition. In particular, this Part addresses the
question why it might matter if certain observed average differences
between men and women have resulted, in part, from evolved, biological
processes rather than solely from social forces shaping gender roles. With
21. See infra text accompanying notes 247–90.
22. Note, however, that the simple conclusion that an attribute of a building, organism, or
doctrine is a “spandrel” does not necessarily imply that it is nonfunctional or dysfunctional.
Some spandrels, including those in the Cathedral of San Marco, are coopted to serve decorative,
useful, or even highly adaptive functions. When a spandrel becomes secondarily adaptive in this
way, Gould labels it an “exaptation.” See Gould, Term and Prototype, supra note 18; cf. Sherry
Ahrentzen, The Space Between the Studs: Feminism and Architecture, 29 SIGNS 179, 194 (2003)
(describing account of the design of the Seven Sisters women’s colleges, where “[c]ollege
planners designed spaces intended to isolate and protect students, but such paternalistic
intentions were subverted by the development of a flourishing student culture in which campus
space was put to varying uses.”). Ironically, in the field of architecture itself, a metaphorical
adaptation became an exaptation by virtue of the lived experience of women students which
defied the paternalistic expectations and intent of the design of their space.
23. In this sense, the analogy to architecture and biology is an imperfect one; the structural
spandrels of a building or an organism are as necessary as the design features that give rise to
their existence. The imperfection of the analogy is, however, to the benefit of the doctrine if it
allows a kind of “genetic engineering” that might not be possible in biology.
332
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
respect to the issue of sexual harassment in particular, it concludes that,
when one looks beyond the rhetoric, there is much common ground
between the proponents of biological and social constructionist positions.
Part IV sets out evolutionary theories as they bear more specifically on
certain typical behavior patterns observed in workplace harassment suits. In
addition, this Part demonstrates that evolutionary analysis can be especially
helpful in understanding these behavior patterns and in analyzing sexual
harassment claims in light of the particular evidentiary hurdles often present
in Title VII discrimination cases premised upon sexual harassment.
Although the federal courts exhibit a nearly obsessive concern with context
in determining whether the facts of any particular case rise to the level of
actionable harassment,24 in fact there are certain readily discernible factual
patterns in sexual harassment cases. Accordingly, this Part describes, in
general terms, several scenarios that tend to recur in the case law. It then
breaks down these behaviors into those that tend to be especially relevant to
evolutionary analysis, and explains how they might be understood in
evolutionary terms. Finally, it demonstrates how the common evidentiary
difficulty presented by the intent (or causation) element of a Title VII
harassment lawsuit might usefully be addressed by an approach that takes
into account some of the insights gained from evolutionary analysis.
Part V examines current Title VII sexual harassment doctrine and argues
that there exist at least three general sets of “legal spandrels”—doctrinal
spaces that have come to be understood as primary but which in fact are
mere byproducts of other, functional aspects of the law. Indeed, an
improper focus on these spandrels often runs counter to the fundamental,
designed purposes of the law, and therefore these elements are not only
nonfunctional but dysfunctional. Part IV thus describes and analyzes three
legal spandrels in Title VII sexual harassment law, and then proposes to refocus the doctrine by incorporating the insights gained from the
evolutionary analysis in the previous Part. In particular, the causation
requirement contained in Section 703 should be presumed satisfied in
harassment cases that fall within one of the three broad behavioral patterns
outlined in Part IV. Specifically, where a plaintiff demonstrates that the
conduct of coworkers or supervisors falls within one of these sex-based
harassment patterns, courts should presume both that the conduct satisfies
24. E.g., Davis v. Monroe County Bd. of Educ., 526 U.S. 629, 677 (1999); Harris v.
Forklift Sys., Inc., 510 U.S. 17, 20 (1993); Meritor Sav. Bank v. Vinson, 477 U.S. 57, 67
(1986); O’Shea v. Yellow Tech. Servs., Inc., 185 F.3d 1093, 1096 (10th Cir. 1999); Adler v.
Wal-Mart Stores, Inc., 144 F.3d 664, 667 (10th Cir. 1998); Quiroz v. Hartgrove Hosp., 1999
U.S. Dist. LEXIS 4595 at *50 (N.D. Ill. Mar. 25, 1999); Robinson v. Jacksonville Shipyards,
Inc., 760 F. Supp. 1486, 1494–95 (M.D. Fla. 1991).
37:0321]
SEXUAL HARASSMENT LAW
333
the “because of . . . sex” requirement, and that the employer knows or
reasonably should know that such harassing conduct amounts to sex-based
discrimination. Accordingly, the employer should be held liable based upon
its discriminatory intent. Furthermore, with respect to application of the
Faragher/Ellerth affirmative defense,25 employers, to be reasonable, must
act to prevent or deter situations likely to lead to these harassing behavior
patterns.
Because courts, legislators, and members of society affected by legal
rules all come to the table with assumptions about human nature in general,
and male and female natures in particular, it is incumbent on legal scholars
to acknowledge and engage emerging scientific understandings rather than
ignore them.26 And, more importantly, if antidiscrimination laws are to
remain vital through a moment of scientific uncertainty and through
ongoing discoveries in the fields of human genetics, development, and
behavior, it is crucial that these laws rest on a normative and doctrinal
foundation that is not built on the shifting sands of any single theory of
causation, whether genetic or environmental.
25. In Burlington Industries, Inc. v. Ellerth, 524 U.S. 742 (1998), and Faragher v. City of
Boca Raton, 524 U.S. 775 (1998), decided on the same day, the Supreme Court held that an
employer in a sexual harassment case may assert an affirmative defense to vicarious liability for
harassment by a supervisor only where there has been no tangible employment action taken. To
prevail on the affirmative defense, the defendant must demonstrate both: “(a) that the employer
exercised reasonable care to prevent and correct promptly any sexually harassing behavior, and
(b) that the plaintiff employee unreasonably failed to take advantage of any preventive or
corrective opportunities provided by the employer or to avoid harm otherwise.” Faragher, 524
U.S. at 807.
26. There has been a tendency in law and other fields to label the most reprehensible of
human behaviors (and even of animal behaviors) as pathological, and to treat them as
anomalous. Evolutionary and biological insights have the potential to reveal that the behaviors
are not necessarily pathological, and thus to affect policy choices for minimizing or punishing
socially harmful behavior. See, e.g., SARAH BLAFFER HRDY, THE WOMAN THAT NEVER
EVOLVED 76–95 (1981) (discussing debate over infanticide in primates, and arguing that such
behavior, though relatively rare, is not pathological and follows definite evolutionarily-rooted
patterns); Katharine K. Baker, What Rape Is and What it Ought Not To Be, 39 JURIMETRICS J.
233, 236 n.15 (1999) (citing studies that demonstrate that many people view rape as caused by
mental illness, and referencing legislative history of Federal Rule of Evidence 413 to the effect
that rapists constitute a “depraved” minority); Jones, Child Abuse, supra note 4, at 1214–16
(exploring the possible legal implications of scientific understandings that grew out of
evolutionary work on infanticide); cf. Owen D. Jones, Realities of Rape: Of Science and
Politics, Causes and Meanings, 86 CORNELL L. REV. 1386, 1401 (2001) [hereinafter Jones,
Realities] (reviewing RANDY THORNHILL & CRAIG T. PALMER, A NATURAL HISTORY OF RAPE:
BIOLOGICAL BASES OF SEXUAL COERCION (2000)) (discussing the widespread, yet erroneous,
belief that sexual coercion and sexual violence are a uniquely human phenomenon).
334
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
II. APPLICATION TO SELECTED CONTEXTS
This Part offers a very brief illustration of how the proposed approach
would apply to some factual scenarios that have often been problematic
under current sexual harassment theory and doctrine. It is not intended to be
comprehensive, but rather to situate and make concrete the evolutionary and
doctrinal discussion that follows in the next several Parts of the Article.
A. Preexisting Sexualized Atmosphere in the Workplace
Jane works for the Big Man Machismo Corporation (BMMC). After she
has been with BMMC for about six months, she is promoted and transferred
to a division that had until that time been all-male. The other employees, as
well as the supervisor, often use crude language and gestures, describe to
one another their sexual exploits, and hang “girlie” pictures in their work
areas. This behavior is identical before and after Jane enters the workplace.
Jane complains about the sexualized atmosphere, but it continues.
Under existing doctrine and theory, courts have some difficulty deciding
whether Jane has been discriminated against under Title VII. She has not
been treated differently than the men at BMMC; furthermore, it does not
appear she has been specifically “treated” at all. She has simply entered a
preexisting work environment and now argues that this environment is
hostile to her because she is a woman. The individual actors involved in
creating the hostile environment (Jane’s coworkers and/or supervisors) have
no apparent bad intent toward Jane; they simply continue to behave as they
have always behaved. In cultivating or participating in a “macho”
atmosphere in their workplace, have Jane’s coworkers “discriminated”
against her “because of” her sex?
Whereas the answer to this question is not always clear under sexual
harassment doctrine as currently understood, an evolutionary analysis aids
in understanding the true role of sex-based causation in this hypothetical
workplace. As described in Part IV, an evolutionary analysis demonstrates a
strong likelihood that (1) the males in this workplace behave as they do both
because they are male and in order to signal their possession of certain
characteristics either to females or, more likely, to other males; and (2)
women in such a workplace are likely to be differentially affected, on
average, by such behavior.
37:0321]
SEXUAL HARASSMENT LAW
335
B. Same-Sex Male on Male Harassment
John works in a warehouse. All of his coworkers are men. Several of the
guys routinely boast of their (hetero)sexual exploits, tell sexual jokes, and
post and comment on “girlie” pictures. John is visibly bothered by these
discussions. In addition, John talks, dresses, and behaves in what some of
the men consider an insufficiently masculine way. After it becomes clear
that John is uncomfortable and is not willing to participate in this
workplace culture, several of his coworkers start calling him “girlie-man,”
grabbing his testicles when he walks by (“bagging” him), poking him in the
buttocks (“goosing” him), and generally escalating the level of macho talk
and behavior whenever he is around.
Under these circumstances, many courts find that John is not entitled to
relief under Title VII because he has not been discriminated against because
of his sex. Though an increasing number of courts are willing to entertain
the theory that John has been subject to “gender norm enforcement” in such
a way as to offend Title VII, others find that this scenario represents
discrimination on the basis of sexual orientation, which is not prohibited by
the statute.
Under the approach suggested in this Article, a court could find that
John’s coworkers likely behaved toward him as they did because he is a
man. In fact, an explanation of the coworkers’ behavior that draws upon the
insights of evolutionary psychology is entirely consistent with feminist
descriptions of the behavior as discriminatory gender norm enforcement. By
treating John in this way, the other men are engaging in male-bonding
behavior and are excluding a male who might not have been sufficiently
aggressive in hunting or war. Their conduct might also be understood as a
form of status and gender signaling. Under prevailing theories of behavioral
biology, this pattern of conduct would tend to be explained as closely tied to
the sex of the target of the behavior.
C. One or a Few Women in a Predominantly Male Workplace
Michele and Susan are firefighters. All of the other employees in their
unit are men. The men seem to dislike Michele and Susan, and they exclude
them from social events and avoid interacting with them at work. The
women’s coworkers work together in many ways that make the unit run
more smoothly; they do not offer this collegial collaboration to the two
women, who are often left to fend for themselves and to figure out on their
own certain “tricks of the trade” that the men teach one another.
336
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
In this scenario, Michele and Susan might believe that they are being
treated differently because they are women. Their male coworkers might
argue that they simply do not like Michele and Susan very much and that
their behavior toward the women reflects not sex discrimination but a
personality conflict. Many courts would agree: absent evidence of
sexualized behavior in the workplace or evidence of discriminatory intent
on the part of the men (for example, sexist statements that reflect a belief
that women should not be firefighters), Michele and Susan likely would fail
to state a claim of sexual harassment under Title VII because they cannot
demonstrate that they have been treated badly “because of . . . sex” as
required under the statute.
Under my approach, this workplace context characterized by small
numbers of women and a traditionally male-identified occupational skill set
should raise a presumption that the differential treatment of the women has
occurred because of sex. Furthermore, regardless of the subjective intent or
motives of their male coworkers whose collective conduct has created the
hostile working environment, the employer would be presumed to have
acted (or failed to act) with the requisite intent under the statute. Based
upon an evolutionary analysis that would predict harassing behavior by
males against females under certain environmental contexts, employers
which are found to have maintained such environments or allowed them to
persist should be held to have acted with the intent required to sustain a
cause of action under Title VII for hostile work environment sexual
harassment.
III. THE DEVELOPMENT WARS: BEYOND GENES VERSUS ENVIRONMENT
It is nearly impossible to read a newspaper or popular news magazine
without finding an account of a skirmish in the nature-nurture wars.27 If
there is a point of agreement about the nature of the human mind, it is the
27. Most recently, the remarks of Laurence Summers, president of Harvard University,
have crystallized the debate over whether there might exist innate differences in ability between
men and women. See, e.g., Sam Dillon and Sara Rimer, No Break in the Storm over Harvard
President’s Words, N.Y. TIMES, Jan. 19, 2005, at A14; See also, e.g., Sharon Begley, The
Nature of Nuture: How the Environment Can Shape Our Genes, WALL ST. J., June 14, 2002, at
B1; Do You Think Your Actions Have Been Predetermined by God?, WASH. POST, Jan. 4, 2004,
at C11; John Pekkanen, Nature or Nurture?, WASHINGTONIAN, Oct. 2000, at 64; Michael
Prowse, Nurture Versus Nature is a Childish Debate, FIN. TIMES, Sept. 1, 2003, at 17; Matt
Ridley, What Makes You Who You Are; Which is Stronger—Nature or Nurture? The Latest
Science Says Genes and Your Experience Interact for Your Whole Life, TIME, June 2, 2003, at
54; Tom Siegfried, Of Mice, Men and Behavior; Scrutiny Reveals Flaws of Nature vs. Nurture
Debate, DALLAS MORNING NEWS, May 12, 2003, at 1E; The Nature/Nurture Debate: ‘Blank’
Slate Born Quite Full, ATLANTA J.-CONST., Sept. 21, 2003, at 3C.
37:0321]
SEXUAL HARASSMENT LAW
337
universal acknowledgment that we are endlessly fascinated by ourselves.
Studies purporting to offer genetic explanations of why humans do the
things that they do, and why they feel the things that they feel, are certain to
garner media attention regardless of the soundness of the underlying
findings.28 Where those studies focus on biological differences between
males and females, the media attention is even more intense.29
28. See, e.g., Genetta M. Adams, Are Genes to Blame for Celeb Gossip?, PITTSBURGH
POST-GAZETTE, July 31, 2004, at B14 (“Our obsession with celebrity gossip and culture may be
part of our genetic code.”); Anjana Ahuja, Our Fateful Genes, TIMES (London), Dec. 4, 2003, at
14 (discussing role genetics plays in determining many aspects of life); Sharon Begley & Erika
Check, Sex and the Single Fly, NEWSWEEK, Aug. 14, 2000, at 44 (discussing promiscuity);
Sandra Blakeslee, How Does the Brain Work?, N.Y. TIMES, Nov. 11, 2003, at F4 (discussing
how genes affect brain activity and other traits); Carey, supra note 3, at F1 (“[T]he urge to
extract a pound of flesh, researchers find, is primed in the genes.”); George Howe Colt, Were
You Born That Way?, LIFE, Apr. 1998, at 38 (discussing the nature versus nurture debate and
media reporting of genetic discoveries); Clive Cookson, Gene Defect Offers Clue to Causes of
Autism, FIN. TIMES, Sept. 10, 2003, at 6; Gene Linked to Shyness, TIMES (London), Apr. 15,
2003, at 6; Erica Goode, Jealous? Maybe it’s Genetic. Maybe Not., N.Y. TIMES, Oct. 8, 2002, at
F1; Dean Hamer, God is in Your Genes, TORONTO STAR, Oct. 2, 2004, at L10 (discussing the
“God gene”); Helping Young Girls Avoid Depression, CONN. POST, Oct. 24, 2003 (discussing
the use of genes to identify young girls at risk for depression); Ronald Kotulak, IQ May be
More than Matter of Genes, CHI. TRIB., Dec. 2, 2003, at C1; Sarah Lesher, A Move Toward
‘Designer Genes’?, THE HILL, June 18, 2003, at 46 (discussing the manipulation of genes for
choice of particular traits); Ian Sample, Mindless Brutality? No, It’s the Spiteful Gene,
GUARDIAN (London), Sept. 3, 2004, at 9 (“gene or set of genes for spitefulness”); Shankar
Vedantam, Desire and DNA: Is Promiscuity Innate?; New Study Sharpens Debate on Men, Sex
and Gender Roles, WASH. POST, Aug. 1, 2003, at A1 (discussing male promiscuity); Ingrid
Wickelgren, Discovery of ‘Gay Gene’ Questioned; Gene Supposedly Responsible for
Homosexuality, SCIENCE, Apr. 23, 1999, at 571. These examples represent a mere fraction of
articles that have appeared in the popular press over the last few years. For an argument that
such media coverage, particularly on the subject of genetic male-female difference, is distorted
and sensationalistic, see Cheryl Brown Travis, Talking Evolution and Selling Difference, in
EVOLUTION, GENDER, AND RAPE 3, 9–17 (Cheryl Brown Travis ed., 2003).
29. One Sunday issue of the New York Times contained several articles on male-female
differences. The cover article of the magazine section, with the headline “Q: Why Don’t More
Women Get to the Top? A: They Choose Not To: Abandoning the Climb and Heading Home,”
suggested that many highly educated professional women leave the workforce not because of
discrimination, but because they have a reasonable excuse to do so. See Lisa Belkin, The OptOut Revolution, N.Y. TIMES, Oct. 26, 2003, § 6 (Magazine) at 42. A front page article in the
business section of the same issue reported and discussed a study finding that parents of girls
are more likely than parents of boys to divorce, and positing that this higher divorce rate might
help explain gender gaps in employment. See David Leonhardt, It’s a Girl! (Will the Economy
Suffer?), N.Y. TIMES, Oct. 26, 2003, section 3, at 1; see also Debra Nussbaum, Education;
Lessons in Diversity And Aggressive Recruiting, N.Y. TIMES, Oct. 26, 2003, section 14NJ, at 6;
David Rohde, India Steps Up Effort to Halt Abortions of Female Fetuses, N.Y. TIMES, Oct. 26,
2003, section 1, at 3.
338
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
The debate over nature versus nurture, and among feminists, behavioral
biologists, and law and evolution theorists, has longstanding roots.30 In its
starkest terms, the nature argument posits that human behavior and
psychology are in large part controlled and determined by genes,31 while the
nurture argument posits that environmental influences are determinative and
controlling.32 Though repeatedly pronounced dead, the nature-nurture
debate refuses to remain buried.33
30. See Anne Fausto-Sterling, et al., Evolutionary Psychology and Darwinian Feminism,
23 FEMINIST STUD. 403, 406–407 (1997) (review essay) (recounting the longstanding dispute
between Darwinians and feminists beginning with Antoinette Brown Blackwell’s 1875 booklength reaction to Darwin’s Descent of Man, and Selection in Relation to Sex); SARAH BLAFFER
HRDY, MOTHER NATURE: MATERNAL INSTINCTS AND HOW THEY SHAPE THE HUMAN SPECIES
20–22 (1999) (same); Zuleyma Tang-Martinez, The Curious Courtship of Sociobiology and
Feminism: A Case of Irreconcilable Differences, in FEMINISM AND EVOLUTIONARY BIOLOGY:
BOUNDARIES, INTERSECTIONS, AND FRONTIERS 116 (Patricia Adair Gowaty ed., 1997)
[hereinafter FEMINISM AND EVOLUTIONARY BIOLOGY] (summarizing feminist criticisms of
sociobiology and evolutionary psychology).
31. This characterization of the “nature” position is in fact a caricature; no mainstream
scientist or academic who posits genetic bases for behavior actually claims that genes are
determinative or controlling. See, e.g., STEVEN PINKER, THE BLANK SLATE: THE MODERN
DENIAL OF HUMAN NATURE 112–13 (2002) (“Neither [Richard] Dawkins nor any other sane
biologist would ever dream of proposing that human behavior is deterministic, as if people must
commit acts of promiscuity, aggression, or selfishness at every opportunity.”); Jones, Child
Abuse, supra note 4, at 1125–26 (“This Article does not suggest that genes ‘determine’ specific
human behavior. They do not.”). In fact, it is the critics of sociobiology and related fields who
paint the position as “determinist” or “reductionist.” See id. at 105–20. I frame the debate in this
manner simply to emphasize the extremes in order to contrast them. Of course, most people on
both sides of the debate take much more nuanced and sophisticated positions. Indeed, most of
those who rely on biological arguments state that the nature-nurture dichotomy is a false one.
E.g., Owen D. Jones, Time-Shifted Rationality and the Law of Law’s Leverage: Behavioral
Economics Meets Behavioral Biology, 95 NW. U. L. REV. 1141, 1168 (2001) [hereinafter Jones,
Rationality] (“Arguing about whether or not a given behavior is the product of genes or culture
is (as is often noted) like arguing about whether the area of a rectangle is the product of its
length or its width.”). In addition, an emerging theory in evolutionary biology, known as the
“Developmental Systems Approach” or DSA, has challenged the traditional framework of even
those biological accounts that recognize the interrelationship of genes and environment. See
Russell Gray, “In the Belly of the Monster”: Feminism, Developmental Systems, and
Evolutionary Explanations, in FEMINISM AND EVOLUTIONARY BIOLOGY, supra note 30, at 385
(discussing implications of DSA); see also MATT RIDLEY, NATURE VIA NURTURE: GENES,
EXPERIENCE AND WHAT MAKES US HUMAN 128–30 (2003) (summarizing history and current
state of the “developmentalist challenge” in biology); MARY JANE WEST-EBERHARD,
DEVELOPMENTAL PLASTICITY AND EVOLUTION (2003); Ani B. Satz, Testing Access: Toward a
Theory of Entitlement to Genetic Testing, at 90–100 (defining genetic determinism and
discussing the implications of genetic determinism and reductionism) (unpublished Ph.D. thesis,
Monash Univsersity) (on file with author).
32. Similarly, this characterization of the “social construction” position does not account
for the spectrum of views held by those generally associated with it. Although there do exist
academics who seem at times to assert that there is no biological component to human nature or
to differences between the sexes, see Judith Lorber, Believing is Seeing: Biology as Ideology, 7
37:0321]
SEXUAL HARASSMENT LAW
339
This Part summarizes the nature-nurture debate as specifically related to
human sex differences, describing both the arguments from nature and the
arguments from nurture as they are used to explain differences between
women and men. It then focuses on the legal scholarly debate and, in
GENDER & SOC’Y 568, 568–71 (1993), most adherents of a social construction perspective
acknowledge some biological component but stress the amplifying role of culture in building
upon small or socially insignificant biological traits or differences, see David A. Strauss,
Biology, Difference, and Gender Discrimination, 41 DEPAUL L. REV. 1007, 1011 n.9 (1992)
(arguing that the assertion that there are biological tendencies toward certain behavioral or
psychological characteristics “is compatible with the position that society has heightened those
tendencies in illegitimate ways”); cf. Janet E. Halley, Sexual Orientation and the Politics of
Biology: A Critique of the Argument from Immutability, 46 STAN. L. REV. 503, 547–553 (1994)
(analyzing several strands of “essentialism” and “constructivism” in the context of the debate
over the nature of sexual orientation). Still others question whether the issue of cause is relevant
at all to the feminist project. See FINEMAN, supra note 15, at 34 (stating that “[e]ven if
differences are considered to be socially rather than biologically constituted, nothing is
resolved.”). Professor Strauss’s position, ironically, is consistent with a point made by one of
the “most extreme proponents of sociobiology,” E. O. Wilson. See EDWARD O. WILSON,
SOCIOBIOLOGY: THE NEW SYNTHESIS 11–13 (1975) [hereinafter WILSON, SOCIOBIOLOGY]
(explaining the concept of the Multiplier Effect, whereby “[a] small evolutionary change in the
behavior pattern of individuals can be amplified into a major social effect by the expanding
upward distribution of the effect into multiple facets of social life.”); Abrams, supra note 4, at
1025 (criticizing the “small but hardy band” of extreme sociobiologists exemplified by E. O.
Wilson for “extrapolat[ing] from the species to the individual and from physical characteristics
to psychological ones”) (quoting CYNTHIA FUCHS EPSTEIN, DECEPTIVE DISTINCTIONS: SEX,
GENDER, AND THE SOCIAL ORDER 47 (1988)); see also MARGARET MEAD, MALE AND FEMALE, A
STUDY OF THE SEXES IN A CHANGING WORLD 7 (1949) (pointing out the universal tendency of
societies to “elaborate the biological division of labour into forms often very remotely related to
the original biological differences that provided the original clues”), quoted in Browne, Seeking
Roots, supra note 4, at 5 n.3). For a debate about the meaning of the term “social construction,”
see generally 41 DEPAUL L. REV. 981–1056 (collection of articles debating role of biology as
relevant to antidiscrimination statutes). In fact, E. O. Wilson’s personal views concerning
human sex difference can appear distinctly moderate. He has stated that he believes that the
evidence on humans shows that:
modest genetic differences exist between the sexes; the behavioral genes
interact with virtually all existing environments to create a noticeable
divergence in early psychological development; and the divergence is almost
always widened in later psychological development by cultural sanctions and
training. Societies can probably cancel the modest genetic differences
entirely by careful planning and training . . . .
E. O. WILSON, ON HUMAN NATURE 129 (1978), quoted in HRDY, supra note 26, at 200–01 n.1.
It might be noted that this statement of biological sex differences by Wilson, the father of
sociobiology, is strikingly congruent with the definition of social construction in Professor
Strauss’s Reply to Professor Epstein in the DePaul Law Review colloquy. David A. Strauss,
Biology, Difference, and Gender Discrimination, 41 DEPAUL L. REV. 1007, 1011 (1992).
33. See generally PINKER, supra note 31 (detailing the long history of the debate over
biology versus environment); RIDLEY, supra note 31 (same); LESLEY ROGERS, SEXING THE
BRAIN (1999) (evaluating scientific evidence concerning biological sex differences in brain
physiology and function).
340
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
particular, on its application in the context of sexual harassment in the
workplace. It ends with an account of more recent work by evolutionary
biologists, some of whom have approached the issues of sexual selection
and sexual dimorphism from a self-consciously feminist perspective and
have made important and widely recognized contributions to scientific
understanding in the field.
A. Arguments from Nature: Natural Selection and Sexual Selection
The scientific field of evolutionary biology looks at species-typical
physical traits, for example the giraffe’s long neck, the lion’s sharp tooth, or
the human’s opposable thumb, and explores the ways in which such
features could have evolved. One of the ways in which many features
evolve is through the process of “Darwinian,” or natural, selection.34 Those
organisms which embody heritable characteristics35 that help them to
survive and reproduce at the expense of similar organisms which lack those
characteristics pass on their genes in greater proportion to future
generations. Those that don’t, don’t. Thus, every living organism is by
definition the end result of a long line of reproductively successful
ancestors. Biologists and anthropologists might disagree over the details of
Darwinian selection,36 but they generally agree that the process of natural
34. Though natural selection is generally thought to account for much evolution and
speciation, other mechanisms of evolution also operate. These include genetic drift and founder
effects. See Victoria L. Sork, Quantitative Genetics, Feminism, and Evolutionary Theories of
Gender Difference, in FEMINISM AND EVOLUTIONARY BIOLOGY, supra note 30, at 86, 104–05.
35. Some recent scholarship in evolutionary biology questions the centrality of genes to
the concept of heritability. Based upon the definition of heritability pronounced by John
Maynard Smith, who is generally considered among the greatest evolutionists of the Twentieth
Century, these scholars have argued that factors other than genes can be properly considered
“heritable.” Maynard Smith defined heritability as information transfer between the generations.
See, e.g., John Maynard Smith, The Concept of Information in Biology, 67 PHIL. SCI. 177, 178–
79 (2000). Accordingly, it has been argued that learning and culture, among other nongenecentered factors, may properly be considered heritable, and that these may contribute to
evolutionary success. See generally Stephen M. Downes, Heredity and Heritability, in STAN.
ENCYCLOPEDIA PHIL. (forthcoming Fall 2004 ed.) (describing debate over centrality of genecentered view
of
heredity and citing numerous sources), available at
http://plato.stanford.edu/archives/fall2004/entries/heredity/ (last visited Feb. 28, 2005).
36. For example, there has been much debate concerning the level at which natural
selection operates. In other words, scientists disagree about the correct “unit” of natural
selection: the gene, the individual organism, the social group, or the species. Biologist Richard
Dawkins is most clearly associated with the theory of gene selection. See RICHARD DAWKINS,
THE SELFISH GENE 38–39 (1986) (“Any gene which behaves in such a way as to increase its
own survival chances in the gene pool at the expense of its alleles will, by definition,
tautologously, tend to survive. The gene is the basic unit of selfishness.”); see also GEORGE C.
WILLIAMS, ADAPTATION AND NATURAL SELECTION: A CRITIQUE OF SOME CURRENT
37:0321]
SEXUAL HARASSMENT LAW
341
selection is the primary engineer that has designed37 the “survival
machines”38 that are alive today. To the extent that evolutionists disagree
about natural selection, that disagreement tends to be concerned with the
appropriate way to determine whether a particular trait is an “adaptation.”39
EVOLUTIONARY THOUGHT 20–56 (1966). The late Harvard Paleontologist Stephen Jay Gould
was the most visible proponent of the theory of group, or species, selection. See, e.g., Stephen
Jay Gould & Elisabeth A. Lloyd, Individuality and Adaptation Across Levels of Selection: How
Shall We Name and Generalize the Unit of Darwinism?, 96 PROC. NAT’L ACAD. SCI. 11904,
11908–09 (1999); Elisabeth A. Lloyd & Stephen Jay Gould, Species Selection on Variability, 90
PROC. NAT’L ACAD. SCI. 595, 595–97 (1993). For an accessible discussion of this debate, see
generally KIM STERELNY, DAWKINS VERSUS GOULD (2001). See also Marcy F. Lawton et al.,
The Mask of Theory and the Face of Nature, in FEMINISM AND EVOLUTIONARY BIOLOGY, supra
note 30, at 63, 68 (arguing that the prevailing paradigm of “selfish individualism” in
evolutionary biology prevented scientists from examining the data and issues surrounding
sociality in species). After the publication of Williams’s critique of group selection theory, the
gene selection model came to dominate the field. However, some recent literature has revived
the theory of group selection. See ELLIOTT SOBER & DAVID SLOAN WILSON, UNTO OTHERS: THE
EVOLUTION AND PSYCHOLOGY OF UNSELFISH BEHAVIOR 132–58 (1998); Robert Boyd & Peter J.
Richerson, Group Beneficial Norms Can Spread Rapidly in a Structured Population, 215 J.
THEOR. BIOL. 287, 287–88 (2002); Elliott Sober & David Sloan Wilson, Reintroducing Group
Selection to the Human and Behavioral Sciences, 17 BEHAV. & BRAIN SCI. 585 (1994); Ted C.
Bergstrom, Evolution and Group Behavior: Individual and Group Selection, 16 J. ECON. PERSP.
67 (2002), available at http://repositories.cdlib.org/ucsbecon/bergstrom/2002A (last visited
Mar. 7, 2005). This more recent group of selectionist literature, however, recognizes other
levels of selection as significant and is more often characterized as “multi-level” selection
theory.
37. Biologist Richard Dawkins uses the phrase “the blind watchmaker” to illustrate both
the nonpurposefulness and the magnificently functional design of natural selection. See
generally RICHARD DAWKINS, THE BLIND WATCHMAKER (1986).
38. The phrase is Dawkins’s. See, e.g., DAWKINS, supra note 36, at 37 (“[Genes] are the
replicators and we are their survival machines.”); WILLIAMS, supra note 36, at 24 (“Socrates’
genes may be with us yet, but not his genotype, because meiosis and recombination destroy
genotypes as surely as death.”); John H. Beckstrom, The Potential Dangers and Benefits of
Introducing Sociobiology to Lawyers, 79 NW. U. L. REV. 1279, 1281 (1984–85) (“A
considerable and continually increasing body of scientific evidence suggests that all living
organisms, including humans, can be thought of as vehicles that carry genetic material from one
generation to the next.”).
39. Evolutionary biologists generally recognize an architectural element of an organism as
a presumptive adaptation when “it solves an adaptive problem with ‘reliability, efficiency, and
economy.’” Leda Cosmides & John Tooby, Evolutionary Psychology: A Primer, Univ. of Cal.
Santa
Barbara
Center
for
Evolutionary
Psychology,
at
http://www.psych.ucsb.edu/research/cep/primer.html (last visited on Mar. 7, 2005) (quoting
WILLIAMS, supra note 36). The late Professor Stephen Jay Gould, however, frequently criticized
sociobiology and evolutionary psychology for a propensity to explain nearly every trait as an
adaptation. He argued that many characteristics of organisms could better be explained as
byproducts of other, adaptive, traits. E.g., Gould & Lewontin, supra note 1.
342
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
Evolutionary psychology40 is a particular application of evolutionary
biology that applies the model of Darwinian selection and adaptation to
explain mental processes and human behaviors.41 Just as the human brain
40. One might note about evolutionary psychology that the field itself is in an
evolutionarily unstable state. It is young. It is changing rapidly, with new discoveries almost
daily being reported. To push the metaphor further, the field is in the “punctuated” stage of a
punctuated equilibrium. This is partly because evolutionary psychology draws from several
diverse but interrelated fields of study. Thus, and most notably, the explosion of data in the
fields of neuroscience and genetics has profoundly affected not only the empirical base of
evolutionary psychology, but the very conceptual foundations on which it rests. Resting legal
proposals upon this shifting foundation is a perilous business at best and one that must, at the
very least, be approached with a degree of humility.
41. The various labels attached to evolutionary theorizing about humans and animals can
be daunting. The field is sometimes called “sociobiology.” The concept of sociobiology, a term
popularized by Harvard Entomologist Edward O. Wilson through his 1975 book of that title,
grew out of the study of social insects and his insight that many of the social behaviors of these
insects and other social species could be explained in part by evolutionary adaptation. See
generally WILSON, SOCIOBIOLOGY, supra note 32 (explaining a theory of evolutionary change).
Modern sociobiologists have generally focused on studies of and evolutionary explanations for
animal behavior, while those studying human behavior split into two separate fields. “The ones
with strong ties to ecology and ethology . . . joined with ecological anthropologists studying
tribal societies and identified themselves as ‘human behavioral ecologists.’ . . . [T]heoretical
developments in this companion field paralleled those in animal sociobiology.” HRDY, supra
note 26, at xix. Others, whose work grew out of the early sociobiological theories, called
themselves “evolutionary psychologists” and have focused on human cognition and behaviors.
See id. Hrdy suggests that this group has continued to rely on outdated assumptions from early
sociobiology, while contemporary sociobiologists and human behavioral ecologists have
incorporated the newer data and insights and thus have a more accurate and robust explanation
of behavior. See id. Some evolutionary biologists are skeptical of evolutionary psychology and
of what they term “pop-sociobiology” engaged in by nonscientists. Certainly, the relative
intellectual accessibility of the basic tenets of sociobiology, combined with their tendency on a
simplistic level to coincide with widespread social beliefs, sometimes results in false or
misleading information. And because the discussion is so politically charged, critics of one
position or the other often assume malign intent and ulterior motives on the part of the
messenger. On the other hand, data and conclusions are sometimes presented as “science” to a
popular audience when mainstream scientists would seriously question that characterization. 42.
See PINKER, supra note 31, at 41 (“One can say that the information-processing activity of
the brain causes the mind, or one can say that it is the mind, but in either case the evidence is
overwhelming that every aspect of our mental lives depends entirely on physiological events in
the tissues of the brain.”); Cosmides & Tooby, supra note 39 (“The brain is a physical system
whose operation is governed solely by the laws of chemistry and physics.”). It has been noted
that one difficulty humans face in understanding their biology is that the very instrument of
understanding is part of that biology, and that instrument (the brain) is sometimes a liar. One
fascinating illustration involves a man whose left and right hemispheres were disconnected
during surgery. In subsequent experiments, his right side could obey signals given to one side of
his brain, and his left side likewise could obey signals, but the two sides could not
“communicate” and thus, to use a metaphor that in this instance is literal rather than figurative,
the left hand did not know what the right hand was doing. However, each side made up rather
convincing stories to explain what the other hand was doing, and the man in question utterly
believed these stories to be true. For a more detailed account of this bizarre natural experiment,
37:0321]
SEXUAL HARASSMENT LAW
343
might be considered the bridge between the physical body and the higher
faculties that we call the “mind,” so the brain is the bridge between the
disciplines of evolutionary biology and evolutionary psychology. The brain
is a physical part of our biological bodies, and thus is arguably an object of
the evolutionist’s scrutiny. But the brain is also the seat of our motives,
emotions, thoughts, dreams, desires, plans, aesthetic sense, and morality. To
the extent that these intangibles of the human spirit are actually the tangible
results of biochemical and bioelectrical processes in the biophysics of the
brain, they are potentially a proper subject for evolutionary explanations.42
The relatively new field of evolutionary psychology takes up the challenge
of proving that certain psychological characteristics are heritable, highly
particularized, and the result of Darwinian selection.43
see PINKER, supra note 31, at 43. Similarly, humans have an illusion of broadly detailed vision
that is actually a lie convincingly created by the brain. There is currently an active conversation
in the philosophical and vision science literature regarding the significance of this illusion
created by the brain. See, e.g., Alva Noë, Is the Visual World a Grand Illusion?, 9 J.
CONSCIOUSNESS STUD. 1, 2–3 (2002), and sources cited therein; cf. Lawton et al., supra note 36,
at 68 (noting Habermas’ argument that “insofar as science is engaged in its own project . . . it
cannot examine the social preconditions of its own existence”). This idea that there is distortion
inherent in the measurement of the thing that performs the measuring is similar to the
Heisenberg uncertainty principle in physics, which posits that the act of measuring changes the
properties of the measured. For a discussion of the principle, see AMIR D ACZEL,
ENTANGLEMENT 73–81 (2001). For an application of the principle to law, see Laurence H.
Tribe, The Curvature of Constitutional Space: What Lawyers Can Learn from Modern Physics,
103 HARV. L. REV. 1 (1989).
43. Evolutionary biologists have long sought to explain the adaptive function of human
intelligence in general. George C. Williams suggested that “advanced mental qualities might
possibly be produced as an incidental effect of selection for the ability to understand and
remember simple verbal instructions early in life. . . . Don’t tease the saber-tooth.” WILLIAMS,
supra note 36, at 15. For a description of the various theories proposed to explain the sudden
and rapid expansion in human brain size some three million years ago, see RIDLEY, supra note
3, at 309–44; see also GEOFFREY MILLER, THE MATING MIND: HOW SEXUAL CHOICE SHAPED
THE EVOLUTION OF HUMAN NATURE 3 (2000) (advancing the theory that many qualities of the
human mind are more persuasively explained by sexual selection than by natural selection).
344
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
Evolutionary psychology thus styles itself “an approach to the
psychological sciences in which principles and results drawn from
evolutionary biology, cognitive science, anthropology, and neuroscience are
integrated with the rest of psychology in order to map44 human nature.”45 As
an approach to understanding the human mind, evolutionary psychology
rests upon certain fundamental and fairly uncontroversial assumptions: that
human emotions and behavior are a result of certain processes in the brain;
that the brain is a part of the biological body; that the body has evolved such
that certain heritable traits which were more adaptive were passed down in
greater quantities and thus came to predominate in the species. Humans are
animals; the genes that exist in the human population represent evolutionary
or Darwinian “success.”46 According to evolutionary psychologists,
behavior is, quite bluntly, nothing more than movement, and movement is
initiated by the brain.47
It is widely accepted that many species of animals have evolved highly
specialized behavior patterns that they regularly exhibit under particular
circumstances.48 Thus, for example, bears hibernate in winter, squirrels bury
nuts, female black widow spiders mate with and then decapitate and
cannibalize their unfortunate reproductive partners. The most fundamental
disagreement between evolutionary psychologists and the Standard Social
44. The “mapping” metaphor is common in discussions of genetics and evolution. See
generally STEVE OLSON, MAPPING HUMAN HISTORY: DISCOVERING THE PAST THROUGH OUR
GENES (2002); Carl T. Hall, Genes Generate a Map; Study Tracks Human Evolution, Migration,
S.F. CHRON., June 9, 2003, at A4; Rosie Mestel, Draft of Chimp Genetic Map Published;
Scientists Place the Outline on a Web Site So It Can Be Compared with the Human Genome,
L.A. TIMES, Dec. 11, 2003, at A24; Raja Mishra, Scientists Hail Completion of Genome Map,
BOSTON GLOBE, Apr. 15, 2003, at A3. The metaphor suggests a process of finding and marking
fixed points on an unchanging field. This implication of the term is probably misleading,
especially as used to refer to as complex and multifarious a concept as “human nature.” Science
writer Matt Ridley, in contrast, has compared the role of genes in building an organism to the
use of a recipe in baking a cake. See RIDLEY, supra note 31, at 34.
45. Leda Cosmides & John Tooby, Evolutionary Psychology and the Emotions, in
HANDBOOK OF EMOTIONS 91 (M. Lewis & J. M. Haviland-Jones eds., 2000), available at
http://www.psych.ucsb.edu/research/cep/emotional.html.
46. The Darwinian notion of “survival of the fittest” is often said to be a tautology. Under
Darwinian logic, those organisms that are present in the world today are the descendants of
reproductively successful ancestors. “Success” is measured simply by existence. For an
argument that this view is overly simplistic and misleading, see PAUL R. EHRLICH, HUMAN
NATURES: GENES, CULTURES, AND THE HUMAN PROSPECT 16–20 (2000).
47. See Cosmides & Tooby, supra note 39. Thus, organisms that do not move do not have
brains, and in fact organisms that do not move for portions of their lifecycles do not have brains
during those portions. Id.
48. See generally WILSON, SOCIOBIOLOGY, supra note 32.
37:0321]
SEXUAL HARASSMENT LAW
345
Science Model (“SSSM”)49 as widely accepted in the social sciences for the
last century or so, turns on the extent to which humans share these kinds of
evolutionarily influenced behaviors with other animals.50 According to the
SSSM, the human mind is a “blank slate” on which culture is free to
inscribe virtually anything. The potentially revolutionary aspect of the
paradigm shift proposed by evolutionary psychologists is the extent to
which the blank document actually comes preloaded with certain evolved
design templates.51
Evolutionary psychology posits that the brain is composed of many
“modules” that have very specific functional purposes that have evolved
because they were adaptive for our ancestors in the environment that
prevailed during most of human evolutionary history.52 Herein lies perhaps
49. Evolutionary psychologists define the SSSM, with which they view their assumptions
in direct conflict, as the idea, prevalent over the past century in psychology, sociology, and
anthropology, that:
all of the specific content of the human mind originally derives from the
‘outside’—from the environment and the social world—and the evolved
architecture of the mind consists solely or predominantly of a small number
of general purpose mechanisms that are content-independent, and which sail
under names such as ‘learning,’ ‘induction,’ ‘intelligence,’ ‘imitation,’
‘rationality,’ ‘the capacity for culture,’ or simply ‘culture.’
Cosmides & Tooby, supra note 39. For a description and critique of the rise of the SSSM, see
PINKER, supra note 31, at 67–69; see also Paul Rubin, The State of Nature and the Evolution of
Political Preferences, 3 AM. L. & ECON. REV. 50, 51 (2001) (describing SSSM and arguing that
this conception of human behavior is “false and in many respects misleading”).
50. This debate traces back to Descartes, who drew a firm line between humans and
animals, and between mind and body. See RIDLEY, supra note 31, at 10–12 (“René Descartes
had decreed firmly in the seventeenth century that people were rational and animals were
automata.”). William James proposed early in the twentieth century that humans actually exhibit
a wider variety of instinctive behaviors than do other animals, but his theory was soon dwarfed
by the behaviorist school of psychology epitomized by B. F. Skinner. See PINKER, supra note
31, at 19–22 (describing history of debate on this issue).
51. The study of linguistics is an illustration of this more general trend. Professor Stephen
Pinker, a language expert at Harvard, has demonstrated in fascinating detail that, although there
is great variation among languages, there is at bottom a highly specific “language instinct” that
is universal in humans. See STEPHEN PINKER, THE LANGUAGE INSTINCT 26–27, 31 (1994);
STEPHEN PINKER, WORDS AND RULES: THE INGREDIENTS OF LANGUAGE 197 (1999). For an
explanation of the theory that the human brain is composed of highly specific “modules” that
are pre-programmed to solve particular problems that faced our ancestors during much of
human prehistory, see Cosmides & Tooby, supra note 39; Gould/Dennett debate, supra notes
17–18 and accompanying text; RIDLEY, supra note 3, at 313–44.
52. Evolutionary theorists dub this the “environment of evolutionary adaptedness,” or
EEA. On a simplistic level, it can be understood as the environment that faced our huntergatherer ancestors for most of the roughly ten million years during which the human brain
evolved. However, the more accurate, and complex, definition of the EEA is “the statistical
composite of selection pressures that caused the design of an adaptation.” Cosmides & Tooby,
supra note 39. In other words, different biological traits may be tied to different EEAs. In
346
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
the most profound disagreement about the findings and hypotheses of
evolutionary psychology: Is the human brain more like a computer with a
fairly simple software program containing a general algorithm that says, in
effect, “solve this problem any way you see fit,” or does it come pre-loaded
with hundreds or thousands of software programs containing detailed
algorithms for various environmental challenges?53 Those who stress the
“nurture” element in the nature-nurture debate favor the former description;
those leaning toward “nature” favor the latter.54
essence, the EEA is convenient shorthand for a concept, not a distinct period in historical or
evolutionary time.
53. The computer analogy has been widely used in the scientific literature, for example,
DAWKINS, supra note 36, at 52 (“Brains may be regarded as analogous in function to
computers.”); Cosmides & Tooby, supra note 39 (“The brain’s function is to process
information. In other words, it is a computer that is made of organic (carbon-based) compounds
rather than silicon chips.”). The more scientists learn about the brain, and the more sophisticated
computer programming becomes, the more apt the analogy seems.
54. As virtually any participant in the debate will quickly insist, this is a false dichotomy.
There is no nature without nurture, and vice versa. See RIDLEY, supra note 31, at 3:
For more than fifty years sane voices have called for an end to the [nature
v. nurture] debate. Nature-versus-nurture has been declared everything from
dead and finished to futile and wrong—a false dichotomy. Everybody with
an ounce of common sense knows that human beings are a product of a
transaction between the two. Yet nobody could stop the argument.
Id.; Browne, Sex and Temperament, supra note 4, at 982 (“It is easy to view the question [of sex
difference] as entailing a choice between ‘nature’ and ‘nurture,’ yet that is a false dichotomy.”);
Cosmides & Tooby, supra note 39, at 15:
Evolutionary psychology is not just another swing of the nature/nurture
pendulum. A defining characteristic of the field is the explicit rejection of the
usual nature/nurture dichotomies—instinct vs. reasoning, innate vs. learned,
biological vs. cultural. What effect the environment will have on an organism
depends critically on the details of its evolved cognitive architecture.
Id. As Stephen Jay Gould has pointed out, however, disclaiming a singular reliance on genes as
opposed to environment hardly ends the debate—it matters little that the genes necessarily
interact with the environment if the proponent continues to claim that the heritable portion of
the trait is fixed, and varies depending on some insidious classification. See STEPHEN JAY
GOULD, THE MISMEASURE OF MAN 34–35 (1981) (discussing author’s debate with Charles
Murray, one of the authors of The Bell Curve, over this issue). Especially relevant in this regard
is the emerging recognition of “epigenetic factors, [which] are the contributions to a cell’s
environment by genes in other cells of the same individual.” WEST-EBERHARD, supra note 31, at
112. An often-used example of the interaction among genes, and between genes and
environment, is the phenotypic expression of genotype in female ants and honeybees. Because
of reproductive features unique to social insects, sister ants are more closely related to one
another genetically than they are to either their parents or their own offspring. These sisters
share, on average, seventy-five percent of their genes. Their phenotypes, however, differ
markedly based solely on particular factors in their environment, including especially what they
are fed. Those individuals fed a normal diet become workers. Those fed a special diet of “royal
jelly” from the queen will themselves become queens and have a drastically different phenotype
from those sisters, not fed the jelly, who are otherwise genetically nearly identical to them. See
OXFORD DICTIONARY OF BIOLOGY 526 (2000) (definition of “royal jelly”); WILSON,
37:0321]
1.
SEXUAL HARASSMENT LAW
347
Sex Difference Through an Evolutionary Lens
According to evolutionary theory, sex differences within species are
largely the result of sexual selection. Darwin understood sexual selection as
a particular application of the principle of natural selection. He viewed male
competition for females as the driving force of sexual selection.55 If larger,
stronger, more aggressive or more dominant males tend to win the contest
for access to reproductive females, then those traits, if heritable, will be
selected and will become widespread throughout the population. Female
choice is part and parcel of this theory: To the extent that females
differentially choose males with certain characteristics, over time those
characteristics will become prevalent in the population.56 According to
Darwin,
when the males and females of any animal have the same general
habits of life, but differ in structure, colour, or ornament, such
differences have been mainly caused by sexual selection; that is,
individual males have had, in successive generations, some slight
advantage over other males, in their weapons, means of defence,
or charms; and have transmitted these advantages to their male
offspring.57
Contemporary biologists, elaborating upon Darwin’s explanation,
distinguish two types of sexual selection. Male-male competition for
females that takes the form of fighting, dominance and status-seeking
behavior, and the like, is referred to as “intrasexual selection.” In contrast,
selection that is more easily understood as taking place in the context of
SOCIOBIOLOGY, supra note 32, at 160; see also WEST-EBERHARD, supra note 31, at 72–77
(discussing polymorphisms resulting from environmental and developmental factors). A recent
experiment that demonstrated a similarly dramatic phenotypic effect involved infant mice,
otherwise genetically similar, whose mothers were fed one of two different diets while pregnant.
Depending on the maternal diet, the mice were born with either brown or white fur. See Sharon
Begley, Diet During Pregnancy Could Have Effects That Last Into Adulthood, WALL ST. J.,
Aug. 22, 2003, at B1, (describing study and discussing “epigenetic changes, defined as those
having no effect on the sequence of molecules that make up a genome,” but which nonetheless
affect the phenotypic expression of a trait); Duke University Medical Center, Common Nutrients
Fed to Pregnant Mice Altered Their Offspring’s Coat Color, SCI. DAILY MAG., Aug. 8, 2003,
available at http://www.sciencedaily.com/releases/2003/08/030801081754.htm.
55. See CHARLES DARWIN, THE DESCENT OF MAN AND SELECTION IN RELATION TO SEX
222 (John Murray ed., 1882) [hereinafter DARWIN, DESCENT OF MAN]; CHARLES DARWIN, THE
ORIGIN OF SPECIES BY MEANS OF NATURAL SELECTION 583 (Penguin Classics 1985) (1859)
[hereinafter DARWIN, ORIGIN OF SPECIES].
56. A heritable genetic trait conferring a 1% reproductive advantage to an individual will,
in only 265 generations, be present in the entire population. See Jones, Rationality, supra note
31, at 1165 n.80 (citing ROBERT TRIVERS, SOCIAL EVOLUTION 28–29 (1985)).
57. DARWIN, ORIGIN OF SPECIES, supra note 55, at 137–38.
348
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
female choice or preference for certain male characteristics is known as
“intersexual selection.”58
It is one of the cardinal tenets of Darwinism that where one sex is gaudy,
the other must be doing the choosing.59 Physiological differences between
males and females60 stem largely from their differing reproductive
58. See ERNST MAYR, WHAT EVOLUTION IS 138 (2001). Some feminist biologists have
noted that intersexual selection resulting from female preference for certain male traits was for
many decades ignored or disparaged by evolutionists because it contradicted closely held beliefs
about the extent of female agency. See Andreas Paul, Sexual Selection and Mate Choice, 23
INT’L J. OF PRIMATOLOGY 877, 878 (2002):
While Darwin’s first mechanism of sexual selection—male-male
competition over access to females—’the law of battle’—was readily
accepted by his contemporaries and scientific peers, virtually none of them
was convinced by the seemingly strange view of females as active, strategic
decision-makers based on a more than dubious aesthetic sense.
Id.; see also Patricia Adair Gowaty, Sexual Natures: How Feminism Changed Evolutionary
Biology, 28 SIGNS: J. OF WOMEN IN CULTURE AND SOCIETY 901, 908–12 (2003).
59. E.g., DARWIN, DESCENT OF MAN, supra note 55, at 214. Humans thus present a puzzle
for Darwinists. The theory posits females as the more choosy sex, yet it is the female who tends
to be the more showy. See RIDLEY, supra note 3, at 134–69. It may be more accurate to say that,
where one sex exhibits a pronounced characteristic that seems to have no particular survival
value (and may, like the peacock’s tail, actually appear to have negative survival value), it is
fairly safe to assume that the trait comes about through the mechanism of sexual selection. In
other words, the trait has evolved either because it has been differentially preferred by the
opposite sex or because it has proved effective in intrasexual competition for reproductive
partners. Under this view, females attempt to enhance their physical attractiveness because
males differentially prefer attractive females. See DAVID M. BUSS, THE EVOLUTION OF DESIRE:
STRATEGIES OF HUMAN MATING 110–14 (revised ed. 2003). Recent work in evolutionary
biology has begun to demonstrate that “choosy males” are more common than the model
formerly supposed. See, e.g., Patricia Adair Gowaty et al., Indiscriminate Females and Choosy
Males: Within- and Between-Species Variation in Drosophila, 57 EVOLUTION 2037, 2037
(2003); Patricia Adair Gowaty et al., Male House Mice Produce Fewer Offspring with Lower
Viability and Poorer Performance When Mated with Females They Do Not Prefer, 65 ANIMAL
BEHAV. 95, 95–96 (2003) [hereinafter Male House Mice]. This recent work suggests that mate
selection is a dialectic process with significant choice exercised by both males and females, and
with significant fitness effects demonstrated when that choice is constrained.
60. Not all species reproduce sexually, and of those that do, not all are divided into two
sexes. The question of why sexual reproduction evolved, and why that form of reproduction has
come to predominate in the higher vertebrate species, is controversial in biology. Sexual
reproduction entails costs, and thus evolutionary biologists have had to advance theories as to
what benefits might outweigh those costs. However, division into two sexes wherein one, the
“female,” provides a relatively large gamete containing more biological material than the other,
the “male,” has been shown through computer simulations to be an evolutionarily stable
strategy that could conceivably have evolved from a nonsexually-reproducing species. See
RIDLEY, supra note 3. In the same vein, speaking of reproductive success as the “goal” of
certain physical or psychological characteristics is in no way meant to imply any conscious
pursuit on the part of the organism. Cf. RICHARD DAWKINS, THE ANCESTOR’S TALE 5 (2004)
(“From our human point of view, the emergence of our remote fish ancestors from water to land
was a momentous step, an evolutionary rite of passage. . . . That is not the way it was at the
37:0321]
SEXUAL HARASSMENT LAW
349
strategies,61 which in turn stem largely from their differing ecological
environments. Thus, for example, in our closest primate relatives, different
environments have resulted in very different reproductive strategies, which
have resulted in differing degrees and kinds of sex differences in each
species.62 It is generally agreed, for example, that the greater the size
differential between males and females of a species, the greater must be the
male intrasexual competition for access to reproductive females.63
More recently, scientists have discovered that relative testicle size in
males of a species seems to be correlated with promiscuity among females.
In species in which females are more promiscuous, males have larger
testicles relative to their body size.64 The evolutionary explanation lies in
time. Those Devonian fish had a living to earn. They were not on a mission to evolve, not on a
quest towards the distant future.”).
61. Of course, “strategy” is used here in the biological sense; there is no normative
implication. A “strategy” is merely a possible way of passing genes onto the next generation. A
“successful” strategy likewise implies only reproductive success. Others have chronicled the
misunderstandings that have resulted from biologists’ use of the word in other contexts. E.g.,
Jones, Realities, supra note 26, at 1404 n.44 (recounting incident in which authors of a recent
book claiming that rape could be understood as a potentially successful reproductive strategy
were heckled off stage and spat upon at a university where they were giving an invited lecture).
62. FRANS DE WAAL & FRANS LANTING, BONOBO: THE FORGOTTEN APE 140 (1997)
(stating that, in determining why chimpanzees and bonobos evolved such distinct social and
sexual behaviors, “ecological conditions were a key factor”); RIDLEY, supra note 3, at 212–17
(explaining the differing strategies of gorillas, chimpanzees, and gibbons in part by different
ecological niches).
63. It is the competition that leads to the size difference, and not the other way around.
Larger males will be better at monopolizing females, and thus will disproportionately pass on
their genes for large size. However, the logic presumably works both ways, and thus this insight
has been used to help scientists determine the human “mating strategy.” See RIDLEY, supra note
31, at 17–23. It is a hallmark of Darwinist writing that cause and effect are often reversed in
order to conform to our normal way of thinking. Scientists understand that this is going on, but
in popular writing the convention can lead to misunderstandings. In The Selfish Gene, Richard
Dawkins frequently adopts purposive or anthropomorphic language conventions, but then
continually deconstructs the language and reminds the reader that there is no “purpose” to
adaptive design and that genes are not “selfish” in the way we normally understand the word.
See, e.g., DAWKINS, supra note 36, at 50:
[I]t is often tedious and unnecessary to keep dragging genes in when we
discuss the behaviour of survival machines. In practice it is usually
convenient, as an approximation, to regard the individual body as an agent
‘trying’ to increase the numbers of all its genes in future generations. I shall
use the language of convenience.
64. The testicles of a male chimpanzee weigh four times as much as those of a male
gorilla, though a chimpanzee is only one-fourth the gorilla’s absolute size. Male gorillas are
twice as large as female; male and female chimps are very close in size. It appears, therefore,
that male gorillas do their competing at the organism level, whereas male chimps do theirs at the
gamete level. See RIDLEY, supra note 3, at 219. For a general discussion of sperm competition
in birds and primates, see Tim R. Birkhead and Peter M. Kappeler, Post-Copulatory Sexual
350
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
the concept of “sperm competition.” According to this theory, larger
testicles produce more sperm; if females are mating with many males, then
those with more sperm swimming for the egg will be at a reproductive
advantage. The male-male competition in these species is not for access to
reproductive females, but for access to the egg.65
What accounts, under evolutionary reasoning, for female promiscuity?
For many years the standard line based on evolutionary psychology was that
males tend to be promiscuous and females tend to be monogamous. This
was “predicted”66 in a ground-breaking paper by biologist Robert Trivers.67
Trivers’ insight was that, in sexually-reproducing species, the sex that
invests more heavily in the offspring will tend to choose quality partners,
whereas the sex that invests less will tend to opt for quantity. A female
mammal, so the reasoning goes, must bear and nurse offspring. Once
pregnant, she cannot reproduce again for a long time. She has a relatively
fixed and limited potential number of offspring; this number does not
increase if she mates with additional males. A male, on the other hand, need
invest nothing but a small amount of time and energy and a few tablespoons
of semen. He can potentially father an enormous number of offspring. Thus,
in adaptationist terms, males who are promiscuous will be more
Selection in Birds and Primates, in SEXUAL SELECTION IN PRIMATES: NEW AND COMPARATIVE
PERSPECTIVES (Peter Kappeler & Carel P. van Schaik eds. 2004).
65. This phenomenon is an especially good illustration of coevolution of males and
females. Evolutionary Biologist Patricia Adair Gowaty has referred to this process as “sexual
dialectics.” See Patricia Adair Gowaty, Sexual Dialectics, Sexual Selection, and Variation in
Reproductive Behavior, in FEMINISM AND EVOLUTIONARY BIOLOGY, supra note 30, at 351, 369–
78.
66. This is not meant to disparage Trivers or his insights, but only to point out that the
fervor with which they were embraced might partly be explained by their snug fit with
preexisting ideas about male nature and female nature. The subsequent refinements in the
theory, which reveal that things are perhaps not as black and white as they first seemed, have
garnered much less media attention. Professor Gowaty notes that parental investment theory has
“achieve[d] axiomatic status in sociobiology and evolutionary psychology during the last thirty
years . . . [in part because] it has enormous intuitive appeal.” Gowaty, supra note 58, at 902.
This is true despite a lack of empirical support for some of the theory’s predictions in “typical”
(i.e. non role-reversed) species, and despite observational data that “often . . . fails to match its
predictions.” Id. Gowaty argues that scientific testing of the theory, along with alternative
hypotheses, is urgently needed. See id. at 903 (“The ‘basic’ natures of males and females remain
to be described in the vast majority of species; interesting alternative hypotheses exist to explain
sex roles; and these alternative hypotheses offer an empirical challenge to those interested in
understanding the causes and consequences of sexual and reproductive behavior.”) (emphasis
added).
67. See Robert L. Trivers, Parental Investment and Sexual Selection, in SEXUAL
SELECTION AND THE DESCENT OF MAN 1871-1971 136 (Bernard Campbell ed., 1972).
37:0321]
SEXUAL HARASSMENT LAW
351
reproductively successful and their “promiscuity” gene68 will spread
throughout the gene pool, whereas choosy females who pick males with
“good genes”69 will be successful and will pass on their choosiness gene.
68. It is very unlikely that single genes are correlated with specific characteristics. This is
simply a convenient shorthand for the notion that whatever heritable genetic material leads an
organism to exhibit a particular trait will be differentially spread throughout the gene pool in
proportion to that organism’s reproductive success. But see Anthony J. Greenberg et al.,
Ecological Adaptation During Incipient Speciation Revealed by Precise Gene Replacement, 302
SCIENCE 1754 (2003) (reporting study in which manipulation of a single gene in a male
population of fruit flies resulted in marked sexual preferences on the part of females).
69. It is unclear exactly what constitutes “good genes” in this scenario. Some writers posit
that females are choosing males who are apt to be good providers and to share in childcare
investment. See, e.g., DAVID M. BUSS, EVOLUTIONARY PSYCHOLOGY: THE NEW SCIENCE OF THE
MIND 99–122 (1999). However, as a theoretical matter this preference on the part of females
arguably is inconsistent with the “promiscuous male/choosy female” scenario and results in an
unstable combination of preferences. See Gillian K. Hadfield, Flirting with Science: Richard
Posner on the Bioeconomics of Sexual Man, 106 HARV. L. REV. 479, 489 n.28 (1992)
(reviewing RICHARD A. POSNER, SEX AND REASON (1992)); see also DAWKINS, supra note 37, at
303 (admitting that his assertion in the first edition that there existed in theory a stable
equilibrium composed of both “coy” and “fast” females and both “philander” and “faithful”
males was incorrect, and that in fact computer simulations had proven that there was no
evolutionarily stable strategy (ESS) composed of these four types); John Maynard Smith,
Evolution and the Theory of Games, 64 AM. SCIENTIST 41 (1976) (applying game theory to
evolutionary behavior and articulating the idea of an “evolutionarily stable strategy,” or ESS).
Others suggest that females would be choosing males on the basis of characteristics that signal
good health, or good reproductive prospects (though this is a somewhat circular argument). See,
e.g., R. A. FISHER, THE GENETICAL THEORY OF NATURAL SELECTION 146–56 (2d ed. 1958)
(“sexy sons” hypothesis); William D. Hamilton and Marlene Zuk, Heritable True Fitness and
Bright Birds: A Role for Parasites?, 218 SCIENCE 384, 384 (1982) (good genes hypothesis);
Gerald S. Wilkinson et al., Male Eye Span in Stalk-Eyed Flies Indicates Genetic Quality by
Meiotic Drive Suppression, 391 NATURE 276, 277–78 (1998) (good genes). This failure to
explain why females have the preferences that they have (e.g. the peacock’s tail; the
nightingale’s song) is a major gap in the Darwinian story. One recent series of experiments
suggests that individuals, both female and male, choose mates who are particularly suited to
them individually. See Jerram L. Brown, A Theory of Mate Choice Based on Heterozygosity, 8
BEHAV. ECOL. 60 (1997); Jeanne A. Zeh & David W. Zeh, The Evolution of Polyandry I:
Intragenomic Conflict and Genetic Incompatibility, 263 PROC. BIOLOGICAL SCI. 1711 (1996).
One theory is that individuals prefer to mate with those with complementary or dissimilar
immune-coating genes, which has the effect of increasing offspring viability. See id. For an
overview and review of the various theories of mate choice and the data supporting each, see
Paul, supra note 58.
352
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
The model of the promiscuous male and the choosy (or “coy”)70 female
was accepted as dogma and trotted out to help explain, among other things,
male infidelity,71 sexual harassment,72 and rape.73 There is much recent data,
however, that does not fit easily with this model and which has demanded
further explanation.74 For example, DNA testing has revealed, to the shock
of many biologists, that supposedly monogamous females of several species
have in fact been massively cuckolding75 their mates.76 In addition, as
previously mentioned, biologists have observed in many primate species
“extraordinarily
aggressive,
anything-but-passive-coy-disinteresteddiscreet” females, as well as males who seemed rather disinterested in these
70. Charles Darwin used this description of female nature, and it is common in writing on
evolutionary biology. See DARWIN, DESCENT OF MAN, supra note 55, at 225 (“The female, on
the other hand, with the rarest exceptions, is less eager than the male. . . . [S]he generally
‘requires to be courted;’ she is coy, and may often be seen endeavouring for a long time to
escape from the male.”) (quoting HUNTER, ESSAYS AND OBSERVATIONS 194 (Owen ed., 1861));
DAWKINS, supra note 36, at 140–165 (discussing intersexual and intrasexual competition using
game theory, and labeling female reproductive strategies “coy” and “fast”).
71. See Cathy Young, Look Who’s Cheating, BOSTON GLOBE, Aug. 11, 2003, at A11
(describing recent cross-cultural study published in the Journal of Personality and Social
Psychology, which found a higher male preference for sexual variety in every society studied, as
relied on by many evolutionary psychologists to show that male promiscuity is biological); see
also Zoe Williams, Comment & Analysis: Sex Cells: The ‘Science’ of Sociobiology Exists Only
to Explain Why Men Are Within Their Rights to Pursue Young Hotties, GUARDIAN (London),
Mar. 4, 2003, at 24 .
72. See Browne, Seeking Roots, supra note 4, at 5.
73. See RANDY THORNHILL & CRAIG T. PALMER, A NATURAL HISTORY OF RAPE:
BIOLOGICAL BASES OF SEXUAL COERCION 35–36 (2000). This is not to imply that the
evolutionary theory was used to justify these social phenomena, though sometimes it was. Most
scholars are quick to point out that description is not prescription, and that explanation has no
normative value. E.g., sources cited infra note 89.
74. In addition, even the original data which seemed to support the model might have been
seriously flawed. In some cases the conclusions of these studies were later shown to have been
biased by the expectations of the observer. For example, instances of female aggression were
ignored, or female promiscuity and nonpassivity were reinterpreted in light of the prevailing
theory. See Lawton et al., supra note 36, at 69–80 (examining data and critiquing explanations
in two ornithological studies).
75. The word is derived, in fact, from the behavior of the female cuckoo, which practices
what is known as “brood parasitism.” Female cuckoos lay eggs that mimic the eggs of other bird
species. They “intimidate or trick the hosts into accepting their eggs” through a variety of
adaptive behaviors. Thus, the host bird invests resources in raising offspring to which it is not
genetically related. For a discussion of brood parasitism in cuckoos and other bird species, see
WILSON, supra note 32, at 364–68.
76. For discussion of these data, see RIDLEY, supra note 3, at 217–19; Lawton et al., supra
note 36, at 73–80. Some recently gathered human data evidence similar patterns. See BUSS,
supra note 59, at 236 (discussing human DNA studies reporting nonpaternity rates averaging
about ten percent).
37:0321]
SEXUAL HARASSMENT LAW
353
sexual solicitations.77 Despite recent refinements in response to new
empirical data, however, the theory of sexual selection remains
scientifically robust and, at least with respect to nonhuman animals, is the
primary scientific model by which sex differences are understood.78
2.
Female Choice in Biology and Evolution—
Emerging Theories
Recent studies of animal behavior in the field of evolutionary biology
highlight the role of female choice and agency in the viability and
reproductive fitness of offspring. Scientists studying animal behavior have
focused on the ways in which constraints on female choice affect
coevolution of male and female characteristics and on the reproductive
consequences of these constraints.79 Experiments have demonstrated that
constraints on female choice can have serious consequences for the wellbeing of females and the fecundity of the group.80 Applied to humans,81
these studies suggest that constraints on female mate choice, whether
through direct sexual coercion or indirect coercion through restriction of
resources, profoundly affect the health and reproductive success of
offspring.
Several behavioral mechanisms by which males constrain female mate
choice and reproduction have been identified by evolutionary biologists.
These include:
77. See Gowaty, supra note 58, at 907–08 (describing empirical observations that conflict
with conventional assumptions about aggressive, promiscuous males and passive, coy females).
78. Indeed, the observational data have led to a recent flowering of research and interest in
female choice as an important mechanism of sexual selection. This research, and current
theories that have grown out of it, are described and discussed infra notes 79–87 and
accompanying text.
79. See HRDY, supra note 30, at 41–42 (discussing the work of evolutionary biologist
Patricia Adair Gowaty and geneticist William Rice in evaluating the possibilities of “free female
choice”).
80. See id. at 41.
81. As a general matter, evolutionary biologists regard such extrapolation as dangerous,
and often criticize those, including evolutionary psychologists, who make leaps from animal
studies to human behavior. See, e.g., Caitilyn Allen, Inextricably Entwined: Politics, Biology,
and Gender-Dimorphic Behavior, in FEMINISM AND EVOLUTIONARY BIOLOGY, supra note 30, at
515; Charles T. Snowdon, The “Nature” of Sex Differences: Myths of Male and Female, in
FEMINISM AND EVOLUTIONARY BIOLOGY, supra note 30, at 276 (noting the great diversity
among primates and cautioning that reasoning from particular species to humans is fraught with
uncertainty). However, I make such a leap here (albeit with some trepidation) because the
results of the animal studies regarding the importance of unconstrained mate choice to
reproductive fitness have been replicated across five diverse species, and thus the case for
generalizing the results seems fairly strong.
354
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
rape or forced copulation, aggressive conditioning of female
behavior, which may be what some of mate guarding is, for
example—behavior exhibited toward females that restrains
females from seeking extrapair fertilizations; priority of access by
males to resources females need for survival and/or reproduction;
the regulation by males of females’ access to resources, so that
females are forced into mating choices they would not otherwise
make.82
When females are “forced into mating choices they would not otherwise
make,” the consequences are serious. In a series of experiments across five
different species, females who were able to mate with their “preferred” male
had offspring with statistically significant increased rates of viability and
reproductive success.83 In addition, the same experiments carried out on
males have demonstrated identical results.84 Furthermore, these studies
demonstrate that the preferred male is different for different females, rather
than having some absolute genetic fitness independent of who is doing the
choosing. All of these results highlight the importance to females, males,
and their offspring of casting a critical eye on both direct and indirect
(economic) constraints on mate choice. Indeed, these studies highlight the
congruence between libertarian, market analyses of preference, feminist
notions of female autonomy, and evolutionary explanations of male and
female behavior.
Finally, evolutionary models of the behavior of social primates strongly
support arguments in favor of increasing the numbers of women in
traditionally male occupations given the goal of decreasing the incidence of
sexual harassment and thereby increasing equality of occupational
opportunity. Studies involving animals have shown that females are much
82. Gowaty, supra note 65, at 366 (citations in quotation omitted) (citing Barbara B.
Smuts & Robert W. Smuts, Male Aggression and Sexual Coercion of Females in Nonhuman
Primates and Other Mammals: Evidence and Theoretical Implications, 22 ADVANCES STUDY
BEHAV. 1–63 (1993); T.H. Clutton-Brock & G. Parker, Punishment in Animal Societies, 373
NATURE 209 (1995); Patricia Adair Gowaty & W.C. Bridges, Behavioral, Demographic, and
Environmental Correlates of Extrapair Fertilizations Uncertain Parentage in Eastern
Bluebirds, Sialia sialis, 2 BEHAV. ECOLOGY 339 (1991)).
83. See Cynthia K. Bluhm & Patricia Adair Gowaty, Social Constraints on Female Mate
Preferences in Mallards’ Anas Platyrhynchos Decrease, Offspring Viability and Mother
Productivity, 68 ANIMAL BEHAV. 977 (2004); Lee C. Drickamer et al., Free Female Mate
Choice in House Mice Affects Reproductive Success and Offspring Viability and Performance,
59 ANIMAL BEHAV. 371 (2000); Patricia Adair Gowaty et al., Mutual Interest Between the Sexes
and Reproductive Success in Drosophlia Pseudoobscura, 56 EVOLUTION 2537 (2002).
84. See Patricia Adair Gowaty et al., Male House Mice, supra note 59, at 95. This latter
result is inconsistent with conventional evolutionary explanations that tend to assume that males
of most species do not engage in significant sexual selection of females, and which tend to
ignore variation among females.
37:0321]
SEXUAL HARASSMENT LAW
355
less vulnerable to harassment and sexual coercion by males when they are
able to form bonds with other females.85 This animal data is supported by
studies in other fields, including anthropology and sociology,86 and it
coincides with feminist legal arguments that focus on the primary
importance of decreasing occupational segregation on the basis of sex.87
B. Arguments from Nurture: Theories of Social Construction
Much of the criticism of evolutionary theories of sex difference has,
predictably, come from feminists. Because evolutionary models have often
been used to explain (and sometimes justify) what many view as oppressive
and subordinating social practices and norms, there has been significant
resistance to any argument that might be characterized as “biological
determinism.”88 Critics of evolutionary explanations of behaviors that
implicate sex and gender roles have tended both to attack the underlying
data and conclusions and to deny the relevance of biology to the behaviors
at issue.89 In contrast to Neo-Darwinist90 behavioral biological theories,
85. See infra sources cited note 236; see also Kamini N. Persaud & Bennett G. Galef, Jr.,
Female Japanese Quail Aggregate to Avoid Sexual Harassment by Conspecific Males: A
Possible Cause of Conspecific Cueing, 65 ANIMAL BEHAV. 89 (2002); Andrea Pilastro et al.,
Female Aggregation and Male Competition Reduce Costs of Sexual Harassment in the
Mosquitofish, 65 ANIMAL BEHAV. 1161 (2003).
86. “Almost without exception, women fare better in a matrilocal system, in which they
stay where they were born, are surrounded by their friends and relations, and have their menfolk
move in, than they do when they must leave home to join their husband’s domain.” Angier,
infra note 141, at 306 (citing sources in endnotes).
87. Vicki Schultz’s work is notable in this regard. See Vicki Schultz, The Sanitized
Workplace, 112 YALE L.J. 2061, 2174–75 (2003) (proposing evidentiary scheme whereby
employers demonstrating increased integration would be favored, and those whose workplaces
were highly segregated by sex would be penalized).
88. Ellen Berscheid, Forward to THE PSYCHOLOGY OF GENDER, at xii (Anne E. Beall &
Robert J. Sternberg eds., 1993) (discussing sociobiology and noting in connection with
biological theories of human behavior that:
[b]iological determinism is sometimes used both as an explanation and as a
justification for the current status and power differences between men and
women. Sociobiology, in fact, has earned its reputation for being the
psychology of sex, violence, and oppression, with the presumed differential
biological bases of behavior being both the reason (“it’s only natural”) and
the excuse (“they can’t help it”) for male domination and violence toward
women.
Id. (citations omitted); see also Gray, supra note 31, at 385. For a recent discussion of the
politics behind the debate over determinism, see Bailey Kuklin, Evolution, Politics and Law, 38
VAL. U. L. REV. 1129 (2004); see also EHRLICH, supra note 46, at 5–7.
89. See Gray, supra note 31, at 385–87 (citing several examples of feminist criticisms of
evolutionary explanations for behavior).
356
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
which seek to explain much behavior and sex difference with reference to
sexual selection and adaptation, social constructionist theories stress the
central importance of culture in shaping behavior. These theories posit that
most human behaviors are learned, and that most differences between men
and women in behavior, preference, cognition, or psychology are created or
greatly magnified by society.91 Indeed, some proponents argue that the
drives or emotions behind behaviors92 are similarly culturally constructed.
With respect to sex role differences, feminist constructionist theories
hold that “[d]oing gender means creating differences between girls and boys
and women and men, differences that are not natural, essential, or
biological. Once the differences have been constructed, they are used to
reinforce the ‘essentialness’ of gender.”93 In this vein, critical legal theorists
in general tend to condemn approaches that they regard as “essentialist.”94
90. The term “neo-Darwinism” is generally used to describe evolutionary theory following
the “modern synthesis” of Mendelian genetics with Darwinism that began in the 1920s. E.g.,
OXFORD DICTIONARY OF BIOLOGY, supra note 54, at 400. See generally EDWARD J. LARSON,
EVOLUTION: THE REMARKABLE HISTORY OF A SCIENTIFIC THEORY (2004).
91. Alternatively, cultural feminists tend to regard the differences as real, but argue that
the male-biased valuation of these various abilities by society creates and reinforces the social,
political, sexual, and economic subordination of women. E.g., CAROL GILLIGAN, IN A
DIFFERENT VOICE: PSYCHOLOGICAL THEORY AND WOMEN’S DEVELOPMENT 5–23 (1982).
92. For example, some social constructionists have suggested that there exist or have
existed societies in which sexual jealousy is unknown. See Stephen K. Sanderson, The
Sociology of Human Sexuality: A Darwinian Alternative to Social Construction and
Postmodernism, unpaginated paper presented at the annual meetings of the American
Sociological
Association
2003;
at
http://www.chss.iup.edu/sociology/Faculty/
Sanderson%Articles, and sources cited therein. In contrast, evolutionary psychologists view
jealousy, and in particular male sexual jealousy, as a “human universal” explainable as an
evolutionary adaptation. See BUSS, supra note 69, at 324–25; Martin Daly & Margo Wilson,
Family Violence: An Evolutionary Psychological Perspective, 8 VA. J. SOC. POL’Y & L. 77,
108–09 (2000).
93. Gray, supra note 31, at 406 (quoting Candace West & Don H. Zimmerman, Doing
Gender, in THE SOCIAL CONSTRUCTION OF GENDER 13, 24 (Judith Lorber & Susan A. Farrell
eds., 1991)).
94. Gender essentialism, under this critique, is “the notion that a unitary, ‘essential’
women’s experience can be isolated and described independently of race, class, sexual
orientation, and other realities of experience.” Angela P. Harris, Race and Essentialism in
Feminist Legal Theory, 42 STAN. L. REV. 581, 585 (1990). Thus, formulations that posit a “male
nature” or “female nature” are essentialist. Several critical theorists have argued that much
feminism is essentialist and has failed to account for the experiences of diverse women. See,
e.g., Kathryn Abrams, Title VII and the Complex Female Subject, 92 MICH. L. REV. 2479,
2482–85 (1994) (arguing that individuals inhabit a subjective identity that is composed of a
fluid mixture of various social categories); Martha Chamallas, Feminist Constructions of
Objectivity: Multiple Perspectives in Sexual and Racial Harassment Litigation, 1 TEX. J.
WOMEN & L. 95, 95 (1992) (“[F]requently what passes for the whole truth is instead a
representation of events from the perspective of those who possess the power to have their
version of reality accepted. The search is on for multiple meanings and multiple
37:0321]
SEXUAL HARASSMENT LAW
357
To speak of a “female nature” that arises from evolutionary adaptation is,
therefore, a form of biological essentialism. Though there may be weaker or
stronger forms of essentialism, depending on how much of a group’s
commonality is presupposed or described,95 all categorization is essentialist
in some sense.96 Accordingly, though the various social constructivisms
perspectives . . . .”); Kimberle Crenshaw, Demarginalizing the Intersection of Race and Sex: A
Black Feminist Critique of Antidiscrimination Doctrine, Feminist Theory and Antiracist
Politics, 1989 U. CHI. LEGAL F. 139; Harris, supra, at 585–86. As Professor Janet Halley has
noted, however, much of the critique of feminist essentialism is more accurately viewed as a
criticism not of essentialism as such, but of racism. See Halley, supra note 32, at 550 n.186
(discussing critical race theory scholarship that criticizes “feminism’s use of the category
‘woman’”). But see Harris, supra, at 585 (making clear she does not mean to accuse Catharine
MacKinnon and Robin West, whose feminist theories she critiques in the article, individually of
racism). Indeed, the feminist movement itself has largely been reaction against the
“androcentric” essentialism of mainstream thought that conceptualized “human” as “male.”
E.g., FINEMAN, supra note 15, at 12–13 (“Men (and therefore, men’s experiences) cannot
continue to be considered the universal norm, the omnipotent rule makers and law givers.
Women’s experiences must also be incorporated into the law, made central to the rules that
govern our lives.”); West, supra note 5, at 2 (arguing that “all of our modern legal theory—by
which I mean ‘liberal legalism’ and ‘critical legal theory’ collectively—is essentially and
irretrievably masculine” because its descriptions and definitions of human value, fear,
aspiration, and harm define and describe the experience of men but not of women). Professor
Abrams’s critique of Title VII essentialism takes the essentialist critique a step further in that
she also focuses on what might be seen as the negative synergies created by the intersection of
several, sometimes seemingly contradictory, identities in a particular individual. In addition to a
straightforward criticism of judicial essentialism as using the white, middle class woman as the
model for Woman, she also points out the problems inherent in the judicial tendency to take an
aggregative approach when courts do recognize the existence of more than one salient trait. See,
e.g., Abrams, supra, at 2497.
95. For an analysis of the “spectrum” of constructivisms, see Halley, supra note 32, at
556–60 (citing Carole S. Vance, Social Construction Theory: Problems in the History of
Sexuality, in HOMOSEXUALITY, WHICH HOMOSEXUALITY? 13, 21 (Dennis Altman et al. eds.,
1989)).
96. Professor Angela Harris, in her critique of the essentialism of feminist legal theory,
notes that “[e]ven a jurisprudence based on multiple consciousness must categorize; without
categorization . . . there can be no moral responsibility or social change.” Harris, supra note 94,
at 586. She argues for more fluidity and instability in categories, and not against all
categorization as such. See id.; see also Linda Hamilton Krieger, The Content of Our
Categories: A Cognitive Bias Approach to Discrimination and Equal Employment Opportunity,
47 STAN. L. REV. 1161 (1995) (relying on social cognition theory understanding of intergroup
bias, which assumes that mental categorization, and cognitive biases that tend to result
therefrom, are inevitable outgrowths of the way that the human mind works); Stephen Jay
Gould, Triumph of a Naturalist, NEW YORK REVIEW OF BOOKS, March 29, 1984, at 58–71
(reviewing EVELYN FOX KELLER, A FEELING FOR THE ORGANISM: THE LIFE AND WORK OF
BARBARA MCCLINTOCK (1983) (“[W]e must categorize and simplify in order to comprehend.
But the reduction of complexity entails a great danger . . . .”), quoted in CYNTHIA FUCHS
EPSTEIN, DECEPTIVE DISTINCTIONS: SEX, GENDER, AND THE SOCIAL ORDER (1988). There are
interesting parallels between what might be thought of as the human psychological “categorical
imperative” (that is, the cognitive necessity of categorization) and the process of classification
358
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
might be more or less essentialist, all share ultimately in elevating culture
over biology in the explanation of gender roles, behavior, and psychology.97
In its starkest form, the social constructionist understanding of gender
difference posits that observed differences do not have any objective or
biological basis98 but rather are created by the dominant power group (men)
in order to perpetuate political, social, and economic inequality and power
hierarchy over the subordinate group (women).99 Although short-term
tactical realities might make it necessary to think of and treat “women” as a
group in order to gain equal political power, or to elevate “feminine”
characteristics and values in order to compensate for their undervaluation by
patriarchal society, “[t]he long-term goal of feminism must be no less than
the eradication of gender as an organizing principle of post-industrial
society.”100
so central to legal analysis. See FINEMAN, supra note 15, at 18–19 (discussing the necessity,
problems, and politics inherent in legal classification both generally and in application to
particular cases). In some sense the human cognitive process is writ large in the legal process.
One might wonder whether the human cognitive biases and defects elucidated by social
cognition theory have analogues in legal categorization processes.
97. One response to the essentialist critique is to recognize that, regardless of the cause of
sex differences in society, these differences do in fact exist and operate in practice to constrain
women’s lives. Martha Fineman, in her articulation of the concept of “a gendered life,” suggests
that the cause of difference is largely irrelevant, and that a focus on the ideological constructs
surrounding sex role differences:
can . . . be useful in forging (temporary) alliances among women across our
differences. All women are, at least to some extent, judged as “Woman”
(according to gendered expectations) in our society. Therefore, we have an
interest in working together on issues such as motherhood, domestic
violence, sexual harassment, and rape that affect us all.
FINEMAN, supra note 15, at 13. This perspective turns the essentialist critique on its head to
suggest that the essentialist construct “Woman” is created and exists in society in general and
that feminists should recognize that reality rather than emphasizing differences between women.
98. “In the social construction perspective, both sex and gender are socially developed
statuses. Biologists and endocrinologists who study hormones now have a much more
complicated picture of ‘sex.’ Female and male sex are no longer seen as two opposite, mutually
exclusive categories.” Judith Lorber & Susan A. Farrell, Principles of Gender Construction, in
THE SOCIAL CONSTRUCTION OF GENDER 7 (Judith Lorber & Susan A. Farrell eds., 1991).
99. See, e.g., CATHARINE A. MACKINNON, FEMINISM UNMODIFIED: DISCOURSES ON LIFE
AND LAW (1987) [hereinafter MACKINNON, FEMINISM UNMODIFIED]; Catharine A. MacKinnon,
Feminism, Marxism, Method, and the State: An Agenda for Theory, 7 SIGNS 515 (1982);
Catharine A. MacKinnon, Feminism, Marxism, Method, and the State: Toward Feminist
Jurisprudence, 8 SIGNS 635 (1983); see also Lorber, supra note 32, at 569 (“I am arguing that
bodies differ in many ways physiologically, but they are completely transformed by social
practices to fit into the salient categories of a society, the most pervasive of which are ‘female’
and ‘male’ and ‘women’ and ‘men.’”).
100. Judith Lorber, Dismantling Noah’s Ark, in THE SOCIAL CONSTRUCTION OF GENDER,
supra note 98, at 355. It should be noted that many feminists do not agree with this statement of
the long-term goal of feminism. For example, some “cultural feminists” in the tradition of Carol
37:0321]
SEXUAL HARASSMENT LAW
359
There is no doubt that social influences play a role in the creation of
gender roles and that differences related to differential adaptive pressures do
not result solely from genes. Though reports of socially egalitarian or even
matriarchal101 societies have never been fully substantiated,102 it is apparent
that there is a diversity of possibility with respect to social organization in
general, and with respect to both the content and status associated with
gender in particular. Though it might be true, as some insist, that patriarchy
(in the sense of gender hierarchy dominated by males) is a “human
universal,”103 it is also the case that some societies are more equal than
Gilligan see in the values expressed by the female voice the key to the future of the human race.
See generally CHARLOTTE PERKINS GILMAN, HERLAND (1979) (describing a fictional utopian
society of women in which social ills are nonexistent); MARLENE ZUK, SEXUAL SELECTIONS:
WHAT WE CAN AND CAN’T LEARN ABOUT SEX FROM ANIMALS 34 (2002) (discussing
ecofeminism, “which draws a connection between ending inequality between the sexes and
solving environmental problems by changing the human relationship to the natural world”);
West, supra note 5, at 72:
Feminism must envision a postpatriarchal world . . . . [I]n a utopian world,
all forms of life will be recognized, respected and honored. A perfect legal
system will protect against harms sustained by all forms of life, and will
recognize life affirming values generated by all forms of being. . . .
Masculine jurisprudence must become humanist jurisprudence, and humanist
jurisprudence must become a jurisprudence unmodified.
Id. Though it is often unclear whether these cultural feminists ground sex differences in biology,
culture, experience, or some combination of these forces, at least some feminist legal scholars
appear to ground women’s distinct value structure upon biology. See id. at 29–36 (explaining
female “connectivity” and nurturance by virtue of the women’s potential or actual experience of
pregnancy and breastfeeding, and explaining female perception of harm and aspiration for
individuation by virtue of women’s actual or potential experience of unwanted pregnancy and
penetration during heterosexual sex).
101. I use the term here as the converse of “patriarchal” as used to describe overall power,
status, and dominance structures and not simply to describe patterns of out-marriage or property
inheritance. Matrilineal societies, though less common than patrilineal, are not unknown.
Sociologist Steven Goldberg argues that every society that has ever been identified has been
characterized by these three institutions: “[P]atriarchy (males fill the overwhelming [number] of
hierarchical positions, . . .) [M]ale [A]ttainment (males attain the high-status roles, whatever
these may be in a given society), and [M]ale [D]ominance (both males and females feel that
[males dominate] in male-female [relations]. . . .).” STEVEN GOLDBERG, WHY MEN RULE: A
THEORY OF MALE DOMINANCE 63 (1993).
102. See generally Steven Goldberg, The Logic of Patriarchy, 17 GENDER ISSUES 53
(1999); Judy L. Ledgerwood, Khmer Kinship: The Matriliny/Matriarchy Myth, 51 J.
ANTHROPOLOGICAL RESEARCH 247 (1995).
103. See STEVEN GOLDBERG, THE INEVITABILITY OF PATRIARCHY 30–31 (1973);
GOLDBERG, supra note 101 at 15–16. There has been a lively scholarly debate over the issue
whether patriarchy as defined by Professor Goldberg is truly inevitable. In 1986, Society
published Goldberg’s then most recent reiteration of his theory alongside the counter-arguments
of several of his critics in a feature entitled Controversies: Patriarchy and Power. There,
Goldberg summarized his view that “not one of the thousands of societies (past and present) on
which we have any sort of evidence lacks any of three institutions: patriarchy, male attainment,
360
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
others.104 If there are and have been some human societies that are vastly
more egalitarian than others, then the arguments about the force of social
construction in shaping gender roles and hierarchies must be at least partly
correct.
In fact, some studies suggest that male and female infants and children
are treated differently by parents, peers, and teachers, making it difficult or
impossible to disentangle cultural from biological determinants of gendered
behavior.105 “Social role theory [holds that] the differences in the behavior
of women and men that are observed in psychological studies of social
behavior and personality originate in the contrasting distributions of men
and women into social roles.”106 Scholars who advance this theory accept
that certain biological differences exist between human males and females;
their disagreement with evolutionary theories of behavioral difference turns
primarily on the degree and kind of these biological differences. The social
role theorists suggest that the male-female divisions of labor in every
and male dominance.” Steven Goldberg, Reaffirming the Obvious, 23 SOCIETY 4, 4 (1986). For
criticisms of Goldberg’s thesis, primarily from feminists, see Susan Abbott et al., Three Voices
in Opposition, 23 SOCIETY 15 (1986) (arguing that differential socialization of the sexes is a
more convincing explanation of patriarchal social institutions than Goldberg’s theory); Cynthia
Fuchs Epstein, Inevitabilities of Prejudice, 23 SOCIETY 7 (1986) (arguing that change, rather
than patriarchy, is the only true societal inevitability, and that change in the hierarchical status
of women is a slow process beset by artificially paternalistic, socially constructed gender roles);
Alice Schlegel, Logic, Gender and Power, 23 SOCIETY 21 (1986) (ultimately conceding that
most societies are in fact patriarchal, but attributing the phenomenon not solely to biology, but
rather to the centrality of women to reproduction in most preindustrial societies).
104. Cf. GEORGE ORWELL, ANIMAL FARM (1945) (some animals are more equal than
others). In general, “relatively egalitarian relations [have tended most often to be] found in
decentralized, nonhierarchical societies with limited technology and especially in simple
economies that derive subsistence from foraging.” Alice H. Eagly et al., Social Role Theory of
Sex Differences and Similarities: A Current Appraisal, in THE DEVELOPMENTAL SOCIAL
PSYCHOLOGY OF GENDER 123, 129 (Thomas Eckes & Hanns M. Trautner eds., 2000) (citing
several anthropological and historical studies); see also Susan Abbott et al., supra note 103, at
19 (arguing that societies may be divided into three levels of gender egalitarianism: clearly male
dominated; “equalitarian,” with men and women sharing essentially equal power; and pseudomale dominant, where socio-cultural ideology honors males, but in reality reflects a female
stronghold of power) (citing generally PEGGY REEVES SANDAY, FEMALE POWER AND MALE
DOMINANCE: ON THE ORIGINS OF SEXUAL INEQUALITY (1981)).
105. For a meta-analysis of the studies of parental treatment conducted through the late
1980s, see Hugh Lytton & David M. Romney, Parents’ Differential Socialization of Boys and
Girls: A Meta-Analysis, 109 PSYCH. BULLETIN 267 (1991). A more recent summary of the
empirical data, from the perspective of social role theory, can be found in Eagly, supra note
104, at 123–25. In general, the studies of parental treatment have found few gender differences
in treatment, with the notable exception that fathers show a measurable difference in their
reactions to and suggestions regarding gender-stereotyped play in younger children. The effect
is greater with respect to fathers’ treatment of sons than of daughters. See Eagly, supra, at 146
(citing the Lytton and Romney study, as well as two 1998 studies that showed similar results).
106. Eagly, supra note 104, at 125.
37:0321]
SEXUAL HARASSMENT LAW
361
known society stem originally from the biological reproductive role of
women (gestation and lactation) and the greater size and upper body
strength of men. However, they challenge the notion that underlying
biological sex differences are psychological or cognitive in nature. Rather,
these scholars argue that “in more complex societies, the physical attributes
of the sexes generally interact with economic and technological
developments to enhance men’s power and status,”107 and that the resulting
social gender roles are reinforced and entrenched by social descriptive and
injunctive norms.108
This school of social psychology thus explains empirical differences in
men’s and women’s behavior as the end result of a complex psychologicalsocial process that involves an initial sexual division of labor into male and
female gender roles, creation of stereotypes based upon that division, and
then conscious and unconscious regulation and reinforcement by both the
individual actor and observers.109 In principle, according to this view,
statistical behavioral and cognitive differences between men and women
might disappear if the sexual division of labor were reduced or
eliminated.110
C. Nature and Nurture in the Law of Sexual Harassment
Evolutionary psychologists, and the legal academics whose work
incorporates evolutionary theories, generally take pains to make clear that in
describing biologically influenced traits, they do not intend to be
prescriptive.111 The so-called “naturalistic fallacy” lies in the tendency to
107. Id. at 129.
108. See id. at 130–36. Descriptive norms are expectations held by members of a given
society “about what people actually do,” whereas injunctive norms are “expectations about what
people ought to do or ideally would do.” Id. at 131 (citing R.B. Cialdini & M.R. Trost, Social
Influence: Social Norms, Conformity, and Compliance, in 2 THE HANDBOOK OF SOCIAL
PSYCHOLOGY 151 (4th ed.) (D.T. Gilbert et al. eds., 1998)).
109. See generally id. at 125–54.
110. See id. at 159–60 (“The demise of most sex differences with increasing gender
equality, a proposition that thus fits popular beliefs about the characteristics of women and men,
is a prediction of social role theory that will be more adequately be tested as more societies
produce conditions of equality or near equality.”).
111. See, e.g., Browne, Biology, Equality, supra note 4, at 654:
A recognition that certain behavioral sex differences have their origins in
biology does not in any way answer the question of whether the differences
are good and to be fostered by society, or bad and to be suppressed. A large
gap may exist between the descriptive is and the prescriptive ought.
Id.; John A. Robertson, Procreative Liberty in the Era of Genomics, 29 AM. J.L. & MED. 439,
451–52 (2003); see also RIDLEY, supra note 3, at 180–81 (“[N]o moral conclusions of any kind
can be drawn from evolution. . . . It is terribly tempting, as human beings, to embrace such an
362
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
equate description with prescription—to assume that the way that
something is suggests the way that it should be.112 Scientists are in the
business of describing the natural world. Policy, they insist, is for others to
determine. Scientists split the atom; politicians drop the bomb.
However, in legal discussions of biology and sex difference, description
and prescription are sometimes conflated, both by those who offer
evolutionary or other biological explanations of difference and by their
critics. Explanations of sex difference, by their nature, tend to employ a
particular methodology that is distinct from that used in other legal analyses
invoking evolutionary explanation of human behavior. This section
describes the law and evolution scholarship on sex difference, and
highlights the assumptions and implications of its evolutionary and
biological approach to explaining human behavior generally and male and
female behavior specifically.
Legal scholars have offered evolutionary explanations for a variety of
“law-relevant” human behaviors113 and for male-female differences in a
number of contexts. Evolutionary theory has been employed to aid in
evolutionary scenario [of asymmetrical parental investment] because it ‘justifies’ a prejudice in
favor of male philandering, or to reject it because it ‘undermines’ the pressure for sexual
equality. But it does neither. It says absolutely nothing about what is right and wrong.”); Margo
A. Bagley, Patent First, Ask Questions Later: Morality and Biotechnology in Patent Law, 45
WM. & MARY L. REV. 469, 479 (2003) (noting the likely operation of the “is-ought fallacy” in
public misunderstanding of the workings of the patent system). The articulation of the
naturalistic fallacy as such is generally attributed to David Hume, for example, Jonathan R.
Cohen, When People Are the Means: Negotiating with Respect, 14 GEO. J. LEGAL ETHICS 739,
775 (2001), but others suggest that it was the philosopher G.E. Moore who first used the term
and systematically criticized “the idea of drawing values from evolution or, for that matter, from
any aspect of observed nature.” ROBERT WRIGHT, THE MORAL ANIMAL: WHY WE ARE THE WAY
WE ARE 330 (1994) (citing Glossup (1967) at 533 for the proposition that the claim that the
fallacy was first identified by Hume is debatable); Jones, Realities, supra note 26, at 1398 n.26.
As Wright points out, John Stuart Mill had earlier made a searing case against the common
tendency to deduce moral “oughts” from nature on the theory that nature is created by God and
therefore that what “is” must reflect divine value. Mill “wrote that nature ‘impales men, breaks
them as if on the wheel, casts them to be devoured by wild beasts, burns them to death, crushes
them with stones like the first Christian martyr, starves them with hunger, freezes them with
cold, poisons them by the quick or slow venom of her exhalations, and has hundreds of other
hideous deaths in reserve.’” WRIGHT, supra, at 331 (quoting John Stuart Mill, “Nature,” in 10
COLLECTED WORKS OF JOHN STUART MILL 385 (J.M. Robson ed., 1969)).
112. Peter M. Cicchino, Building on Foundational Myths: Feminism and the Recovery of
“Human Nature”: A Response to Martha Fineman, 8 AM. U. J. GENDER SOC. POL’Y & L. 73,
76–77 & n.9 (2000) (citing DAVID HUME, A TREATISE ON HUMAN NATURE I, iv, 6 (1904);
WILLIAM K. FRANKENA, ETHICS 99–100 (2d ed. 1973)).
113. Professor Owen Jones has used this phrase when describing the need for a robust
theory of human behavior in order that law can most effectively influence such behavior in
whatever direction desired by policymakers. See, e.g., Jones, Rape, supra note 4, at 833.
37:0321]
SEXUAL HARASSMENT LAW
363
fashioning legal analyses related to child abuse,114 rape,115 certain perceived
irrationalities in behaviors under economic models,116 wage and gender gaps
in employment,117 the “glass ceiling,”118 workplace sexual harassment,119
and the issues surrounding women in combat.120 Scholarship in other legal
areas, as diverse as First Amendment theory and the basis of morality, has
recently seen arguments based on evolutionary models.121 Some of these
explanations rely on the differential adaptive pressures caused by sexual
selection. Thus, in the context of Title VII—a law that expressly prohibits
sex discrimination in the workplace—explanations of behavior that are
based on a theory of sexual selection have obvious relevance.
Evolutionary explanations can be divided into two broad categories for
purposes of assessing their contributions to legal policy arguments.122 First,
114. See Jones, Child Abuse, supra note 4 at 1231–36.
115. See Jones, Rape, supra note 4, at 838–72; see also THORNHILL & PALMER, supra note
73.
116. See Owen D. Jones, The Evolution of Irrationality, 41 JURIMETRICS J. 289, 304–08
(2001) [hereinafter Jones, Irrationality]; Jones, Rationality, supra note 31, at 1190–96.
117. See Richard A. Epstein, The Varieties of Self-Interest, 8 SOC. PHIL. & POL’Y 102
(1990) (arguing that sociobiological theories of sex difference predict that men and women will
specialize in certain “sex roles within the firm,” and that certain asymmetries will emerge in
labor and capital markets); Browne, Sex and Temperament, supra note 4, at 1075–83; Browne,
Seeking Roots, supra note 4, at 74.
118. See Browne, Sex and Temperament, supra note 4, at 1064–75.
119. See Browne, Seeking Roots, supra note 4, at 9–12; Fremling & Posner, supra note 4.
120. See Browne, Women at War, supra note 4, at 58–64.
121. See Jeffrey Evans Stake, Are We Buyers or Hosts? A Memetic Approach to the First
Amendment, 52 ALA. L. REV. 1213 (2001); Theodore P. Seto, A General Theory of Normativity,
Loyola Law School (Los Angeles) Public Law Research Paper No. 2003-26 (Oct. 2003), at
http://ssrn.com/abstract=458261 (last visited Feb. 5, 2005).
122. A third category of evolutionary argument in law, qualitatively distinct from the
others, consists of the use of the concept of evolution as a metaphor or model for legal change
over time. See, e.g., Roberta C. Clark, The Interdisciplinary Study of Legal Evolution, 90 YALE
L.J. 1238, 1242–56 (1981); E. Donald Elliott, Law and Biology: The New Synthesis?, 41 ST.
LOUIS U. L.J. 595, 600–604 (1997) (stating that “the claim is that any system that exhibits the
three features of reproduction, variation and selection by the environment will evolve in the
direction of greater fit with its environment” and that the law as an “open system” satisfies these
criteria); E. Donald Elliott, The Evolutionary Tradition in Jurisprudence, 85 COLUM. L. REV.
38, 55–90 (1985); Herbert Hovenkamp, Evolutionary Models in Jurisprudence, 64 TEX. L. REV.
645, 651–56 (1985); Paul H. Rubin, Why Is the Common Law Efficient?, 6 J. LEGAL STUD. 51,
55–57 (1977) (arguing that legal rules that are not generally accepted in a community are more
likely to be repeatedly litigated and eventually discarded or modified); Jeffrey Evans Stake,
Pushing Evolutionary Analysis of Law or Evolving Law: Design Without a Designer, 53 FLA. L.
REV. 875, 877 (2001) (“If we can see law as an organically developing set of legal ideas, we can
apply evolutionary theory directly.”). Oliver Wendell Holmes often used evolutionary metaphor
when describing the development of law. See generally OLIVER WENDALL HOLMES, JR., THE
COMMON LAW (1881); Oliver Wendell Holmes, Law in Science and Science in Law, 12 HARV.
L. REV. 443 (1899). Though the metaphor is often apt and though I employ it myself through
364
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
there are those theories that attempt to explain physical or behavioral
features common to all members of a particular species. Thus, the giraffe’s
long neck can be understood as an adaptation that was so useful to those
individuals who possessed it that, in relatively short measure,123 it came to
be a giraffe universal. Evolutionary psychologists and behavioral biologists
likewise focus on “human universals.”124 That there is some disagreement
about the true universality of certain traits should not obscure the fact that
there are some things that are universal in our species.125 Law and evolution
arguments that posit the ultimate reasons for these species-typical behaviors
are potentially useful in fashioning effective legal levers with which to
regulate them.126 This method of evolutionary analysis does not supply a
goal; rather, it looks to norms or policies supplied from an outside source
(the legislature, for example), and then examines evolutionary explanations
of behavior in order to explore their potential usefulness in shaping such
behavior to further the given goal.127
Gould’s concept of evolutionary spandrels, the evolutionary analysis in law with which I am
primarily concerned in this Part is not metaphorical. It is, rather, the reliance on evolutionary
explanations for various human traits as a basis for making concrete legal or policy proposals.
123. As noted supra note 56, a trait that provides a relatively small advantage will, all else
being equal, come to predominate in a population in fairly rapid measure.
124. See DONALD E. BROWN, HUMAN UNIVERSALS 130–41 (1991). Some evolutionary
biologists criticize what they regard as the “gene for” arguments of the evolutionary
psychologists on the ground that there are not enough genes in the human genome to account
for all of the universals that have been catalogued.
125. For example, basic anatomical structures (four-chamber hearts, large brains), drives
(thirst, hunger), emotions (anger, love, jealousy), and social interaction. According to Steven
Pinker, there exists:
[A]n astonishingly detailed set of aptitudes and tastes that all cultures have in
common. This shared way of thinking, feeling, and living makes us look like
a single tribe, which the anthropologist Donald Brown has called the
Universal People . . . . Hundreds of traits, from fear of snakes to logical
operators, from romantic love to humorous insults, from poetry to food
taboos, from exchange of goods to mourning the dead, can be found in every
society ever documented.
PINKER, supra note 31, at 55; see also id. at 435–39 (reprinting Donald E. Brown’s List of
Human Universals). Even biologists suspicious of genetic explanations for human behavior and
critical of “genetic determinism” concede that certain broad features of human nature are shared
generally in the species. See EHRLICH, supra note 46, at 12.
126. See Jones, Rationality, supra note 31 (invoking metaphor of a fulcrum and lever to
explain the possibilities of evolutionary explanations of law-relevant behavior).
127. A related but distinct use of evolutionary models in legal scholarship has been to
complement and reinforce theoretical models of law and/or human behavior. In the context of
law and economics, evolutionary psychology has been the basis of the suggestion that perceived
irrationalities in preference or behavior are actually predictable patterns that one would expect
to find based on Darwinian theory. See Jones, Irrationality, supra note 116; Jones, Rationality,
supra note 31. It has been noted that economic and evolutionary approaches are sympathetic
and complementary to one another. See Katharine K. Baker, Gender, Genes, and Choice: A
37:0321]
SEXUAL HARASSMENT LAW
365
For example, evolutionary theory posits that individuals will treat kin
and nonkin differently, and will do so in nonrandom ways.128 This insight
from evolutionary biology has implications for laws related to child abuse129
and inheritance,130 among others. Similarly, evolutionary theory suggests
hypotheses about the ecological and social conditions that are likely to lead
to sexual aggression and coercion.131 Some of these hypotheses are either
counterintuitive or not readily apparent and arguably have been born out by
the data in some studies of sexual aggression and rape in humans.132 Law
and evolution scholarship analyzing these legal issues in light of
evolutionary explanations of the relevant behavior patterns takes a social
policy goal generated from outside the realm of biology (protect children
from abuse; reduce the incidence of rape), and asks how law and policy
might better be structured to reach that goal in light of behavioral insights
gleaned from evolutionary theory.
The proposition that certain broad patterns of behavior might be
grounded in genetics or biology has several implications for law and policy.
Though the point is often lost in the translation from scientific paper to
newspaper, evolutionary biologists do not believe that geneticallyinfluenced behaviors are rigidly determined by genes.133 Rather, expression
Comparative Look at Feminism, Evolution, and Economics, 80 N.C. L. REV. 465 (2002) (noting
that the methodologies and assumptions of the two approaches are similar). It is not unusual to
find law and economics scholars relying on evolutionary theory. E.g., POSNER, supra note 4;
Epstein, supra note 117, at 102; Fremling & Posner, supra note 4; Rubin, supra note 49.
128. The theory of “kin selection” is generally associated with William Hamilton of
Oxford. It holds “that much of animal cooperation and altruism is explained by the success of
genes that cause animals to look after close relatives because they share many of the same
genes.” RIDLEY, supra note 3, at 77. See William D. Hamilton, The Genetical Evolution of
Social Behavior (I & II), 7 J. THEORETICAL BIOL. 1 (1964).
129. See Jones, Child Abuse, supra note 4.
130. See JOHN H. BECKSTROM, SOCIOBIOLOGY AND THE LAW: THE BIOLOGY OF ALTRUISM
IN THE COURTROOM OF THE FUTURE 58–59 (1985); Jeffrey E. Stake, Darwin, Donations, and the
Illusion of Dead Hand Control, 64 TUL. L. REV. 705 (1990).
131. See THORNHILL & PALMER, supra note 73; Margo Wilson & Martin Daly, The Man
Who Mistook His Wife for a Chattel, in THE ADAPTED MIND: EVOLUTIONARY PSYCHOLOGY AND
THE GENERATION OF CULTURE 289 (Jerome H. Barkow et al. eds., 1992); Jones, Rape, supra
note 4, at 909–30. But see EVOLUTION, GENDER, AND RAPE (Cheryl Brown Travis ed., 2003)
(collection of scholarly papers critical of the methodology and conclusions of Thornhill &
Palmer’s book).
132. For example, evolutionary psychological theories of rape causation would predict that
“[t]he ages of victims of attempted and completed rape will be overwhelmingly concentrated
into the part of the female lifespan that is reproductive.” Jones, Rape, supra note 4, at 865. This
prediction is, in fact, born out by the data (contrary to the received wisdom, which holds that
women and girls of all ages are equally likely to be victimized by sexual violence). Id. at 865–
67.
133. Neither do they believe that the predispositions cannot be overridden by free will, even
under circumstances in which they might be said to be “triggered.” The evolutionary argument
366
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
of the behavioral tendencies is closely tied to ecological conditions under
which the behavior might have been adaptive in the so-called “environment
of evolutionary adaptedness,” or EEA.134 Thus, a tendency toward
promiscuity would have been more likely to be triggered in a context in
which promiscuity would result in greater average reproductive success
than would long-term pairing. In the context of rape law, for example, the
suggestion that many males might have behavioral predispositions toward
sexual aggression and coercion under certain ecological circumstances
implies specific avenues of prevention.135 The potential profit of the
evolutionary explanation over the simple empirical observation lies in the
ability of the former to predict the circumstances under which these socially
harmful behaviors are more likely to occur and thus the potential to
structure law and social policy in ways that could have greater impact on
the behavior.
does not imply an absence of free will; rather, the model speaks only of average patterns and of
probabilities. These arguments do suggest that on average it will be more difficult, and thus
more costly, to prevent certain behaviors under circumstances in which they are likely to
emerge. E.g., Epstein, supra note 117, at 113 (“A rule that mandated an equal number or
fraction of men or women in particular trades or professions . . . cannot be achieved without
some substantial social costs.”); Jones, Rationality, supra note 31, at 1145 (“[The] law of law’s
leverage can therefore afford us new, coherent, and systematic power in predicting the
comparative costs to society of attempting to change behaviors through legal means.”); Douglas
A. Terry, Don’t Forget About Reciprocal Altruism: Critical Review of the Evolutionary
Jurisprudence Movement, 34 CONN. L. REV. 477, 485–87 (2002) (“The import of Beckstrom’s
conclusion is that laws may become too costly, and thus undesirable, when they prescribe
behavior that is contrary to natural human behavior.”); see also MARGARET GRUTER, LAW AND
THE MIND: BIOLOGICAL ORIGINS OF HUMAN BEHAVIOR 6 (1991) (“Rules that are at crosspurposes with strong biological predispositions . . . are either short-lived, disregarded, or
both.”). However, it has been suggested that the deeper implications of behavioral biology upon
human autonomy and free will have not thus far sufficiently been explored. See Steven
Goldberg, Evolutionary Biology Meets Determinism: Learning from Philosophy, Freud, and
Spinoza, 53 FLA. L. REV. 893, 895–96 (2001), which states:
The starting point for many discussions of morality is that for people to be
held morally responsible they must have free will. To put it another way, if
determinism is true, that is, if human choices are the result of antecedent
causes, then morality is impossible. . . . Nor can free will be rescued by
saying that human behavior is caused by multiple factors. Thus, the fact that
Jones does not attribute all human behavior to evolutionary biology does not
take determinism off the table. If my choices are caused by subatomic
particles, the environment, the weather, my upbringing, and evolutionary
biology, those choices, however impossible to predict or understand, are still
not free, and it will be difficult to hold me morally accountable for them.
Id.
134. For an explanation of the term EEA, see supra note 52.
135. THORNHILL & PALMER, supra note 73; Jones, Rape, supra note 4.
37:0321]
SEXUAL HARASSMENT LAW
367
The above examples employ evolutionary theory to help explain general
patterns of behavior irrespective of average differences between the sexes in
the expression of those behaviors.136 These scholars argue that the
evolutionary piece of the puzzle of human behavior heretofore has been
neglected in legal analysis and that it should be considered as one among
many useful models that decision-makers might employ in fashioning and
applying legal rules regulating behavior.137
When it comes to describing and explaining differences between the
sexes, as noted above, another kind of human universal is often invoked: the
universal of difference. Thus, for example, Professor Kingsley Browne has
noted that “[t]he sexual division of labor is a human universal.”138 Much of
136. Though rape patterns demonstrate a marked sex difference in behavior patterns, the
evolutionary analyses of these patterns are not concerned with differences per se, but rather with
how best to shape ecological conditions so as to prevent the expression of the behavior.
Although sexual selection and differential male and female reproductive strategies are a part of
the explanation for the behavior, the focus of the analysis is on using the insights gained through
evolutionary analysis to prevent or deter the (already determined to be) socially harmful
behavior rather than to support normative arguments regarding what the appropriate social
policy should be.
137. E.g., Jones, Law and Biology, supra note 4, at 206 (arguing that laws, to be effective,
should “be leveraging against an integrated model of human behavior . . . that . . . reflects the
most complete understanding available of the multiple and complex influences on behavior . . .
incorporating the most rigorously tested developments of the behavioral sciences with the most
careful empirical observations of the social sciences”).
138. Browne, Seeking Roots, supra note 4, at 5. According to classical evolutionary theory,
the sexual division of labor stems originally from the phenomenon of “anisogamy,” or the
difference between the male and female gametes (sex cells). See, e.g., Trivers, supra note 67, at
144 (“The parental investment pattern that today governs the operation of sexual selection
apparently resulted from an evolutionarily very early differentiation into relatively immobile sex
cells (eggs) fertilized by mobile ones (spermatozoa).”). Because the female gamete (the egg) is
relatively large and relatively expensive for the organism to produce, the female tends to be
“choosy” about who will fertilize it. The male gamete (the sperm), in contrast, is relatively tiny
and cheap to produce because it contains no extra energy in the form of cytoplasmic nutrients.
The male, therefore, need not be choosy and in fact has an incentive to be relatively sexually
indiscriminate. This differential investment in gametes, and in mammals the further differentials
entailed by internal gestation and maternal nursing of infants, is the original division of labor. It
leads, in turn, to further divisions based on the male’s relative mobility and the female’s relative
immobility due to the demands of pregnancy and nursing. These differences in parental
investment, which both lead to and combine with the sharp asymmetry in the potential number
of offspring an individual of either sex could have, produce sexual dimorphism in behavior
according to Darwinian theory. See, e.g., David M. Buss, Sex Differences in Human Mate
Preferences: Evolutionary Hypotheses Tested in 37 Cultures, 12 BEHAV. & BRAIN SCI. 1, 1
(1989). In recent years, several “Darwinian feminists” have questioned some of the premises of
Trivers’s theory of differential parental investment in general, and also as applied to humans
and other primates in particular. See generally Gowaty, supra note 58 (describing history of
parental investment theory and citing numerous critiques and refinements of Trivers’s original
ideas).
368
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
the legal argumentation grounded in, or informed by, evolutionary theory
focuses on this asserted human universal of male-female difference. Indeed,
because the question of biological sex difference is central both to sex
discrimination law and to evolutionary biology, it is hardly surprising that
legal arguments grounded in evolutionary biology and psychology would be
focused on this area of law.
1.
Sex Differences in Cognitive Ability and Preferences
Evolutionary psychology posits that average differences between men
and women in cognitive abilities and preferences will result from sexual
selection. By imagining the environment in which humans and their primate
and mammal ancestors lived for most of their evolutionary history,
evolutionary psychologists predict the kinds of traits that would have been
adaptive in males as opposed to females, based in large part on the different
reproductive strategies of males and females.139 They then “test” these
hypotheses against empirical data,140 including observed animal behaviors
(especially of those species most closely related to humans) and measurable
human behaviors.141 If the data match the predictions, a difference is
considered an evolutionary adaptation.142
139. The difference in male and female reproductive strategy is based primarily on
Trivers’s theory of differential parental investment. See supra note 67 and accompanying text;
infra notes 143, 144 and accompanying text.
140. It has been noted that behaviors leave no fossils, so speculation about human behavior
in pre-historical times is just that—speculation. See generally ROGERS, supra note 33 (critiquing
sociobiology on this ground, among others). See also Gould & Lewontin, supra note 1, at 583–
89 (criticizing mainstream evolutionary biologists for their tendency to ignore, disregard, or
discard data inconsistent with the adaptationist explanation for a particular trait).
141. Human societies that are most closely analogous to those assumed to have existed in
the EEA are especially studied in this regard. Thus, sex role differences in hunter-gatherer
societies are particularly emphasized. Of course, data from traditional hunter-gatherer societies
is neither easy to gather nor especially abundant, and it is often subject to serious
methodological limitations. Evolutionary psychologists therefore tend to rely heavily on selfreported data, including surveys, and studies of insular populations such as university students.
The dangers of extrapolating from such data have been well canvassed. See, e.g., NATALIE
ANGIER, WOMAN: AN INTIMATE GEOGRAPHY 322–54 (1999).
142. The question of when a trait should be considered an adaptation is one of the most
controversial in biology. In his classic book ADAPTATION AND NATURAL SELECTION, biologist
George C. Williams suggested that when “a presumed function is served with sufficient
precision, economy, [and] efficiency” so as to “rule out pure chance as an adequate
explanation,” this should be viewed as at least prima facie evidence that the trait is an
evolutionary adaptation. WILLIAMS, supra note 36, at 9–10. Stephen Jay Gould has criticized
evolutionary psychology for seeing adaptation in every behavioral and psychological trait. See,
e.g., Gould & Lewontin, supra note 1. Scientists generally prefer explanations that are
“parsimonious,” that is to say they prefer simpler to more complex explanations. Evolutionary
37:0321]
SEXUAL HARASSMENT LAW
369
There is a growing body of evidence demonstrating that there exist
measurable, though not large,143 statistical differences between men and
women in certain areas of cognition, preference, and ability.144 These
statistical sex differences have been relied on by some legal scholars
employing a particular genre of evolutionary analysis in law. Professors
Kingsley Browne and Richard Epstein, among others, suggest that men and
women exhibit certain measurable average differences in cognition.145
These differences, they argue, can (and should) be understood as
adaptations resulting from the process of sexual selection. For example,
some studies have revealed statistical differences between women and men
psychologists assert that their explanations of many human sex differences satisfy this criterion.
See, e.g., BUSS, supra note 59, at 211 (“Given the power of sexual selection . . . it would be
astonishing to find that men and women were psychologically identical in aspects of mating
about which they have faced different problems of reproduction for millions of years.”).
143. Though the absolute differences are very small, and though there is much greater
within-sex than between-sex variation in all of the studied traits, Professor Browne
demonstrates that, for certain sex-differentiated traits, when the male mean is higher and male
variability is greater (which it is for many of these traits), the result is a much larger percentage
of males than females at the far right of the bell curve. See Browne, Sex and Temperament,
supra note 4, at 1016 n.256 (offering statistical explanation).
144. For a detailed discussion of differences in cognitive abilities and an overview of recent
studies, see BROWNE, BIOLOGY AT WORK, supra note 4, at 25–32 and sources cited therein
(offering examples of average differences suggested by these authors, and citing studies). Note,
however, that biology predicts (and research in fact finds) that the vast bulk of human
intellectual ability shows no statistical difference between the sexes. For arguments that these
findings of sex difference in cognitive ability are flawed, see ANGIER, supra note 141, at 322–
67; CYNTHIA FUCHS EPSTEIN, DECEPTIVE DISTINCTIONS: SEX, GENDER, AND THE SOCIAL ORDER
52–56 (1988); ANNE FAUSTO-STERLING, SEXING THE BODY 3–5 (2000); CAROL TAVRIS, THE
MISMEASURE OF WOMAN 43–56 (1992); Mary Anne Case, Of Richard Epstein and Other
Radical Feminists, 18 HARV. J.L. & PUB. POL’Y 369, 388 n.73 (1995) (“To catalogue all the
errors Epstein makes in his discussion of sociobiology, I would need to be far better trained in
the natural sciences and to write a book.”); Strauss, supra note 32, at 1008 n.2 (“I do not address
Professor Epstein’s claims about sociobiology and neurophysiology, other than to note that the
sociobiological models he relied upon, at least, have been tellingly criticized.”).
145. See, e.g., BROWNE, BIOLOGY AT WORK, supra note 4, at 25 (“The sexes differ in
performance on many cognitive tasks, although they differ little, if any, in general cognitive
ability. Many tests of spatial ability, especially mental rotation, show a consistent male
advantage, while others, such as object location, show a female advantage.”); Richard A.
Epstein, Gender Is for Nouns, 41 DEPAUL L. REV. 981, 988–89 (1992) [hereinafter Epstein,
Nouns] (describing “incontrovertible” evidence that “[t]he division of functions across different
parts of the brain are not the same for men and for women, which in part accounts for why men,
for example, have superior skills in dealing with spacial relations, and women superior language
skills”); Epstein, Two Challenges, supra note 4, at 338. (“The differences between men and
women, then, are not simply matters of size, or even matters of size and strength . . . . They are
also matters of psychology and behavior” that include “giv[ing] directions in different ways”
and “interact[ing] with computers in fundamentally different ways.”).
370
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
in certain spatial abilities.146 Evolutionary psychologists suggest that men,
who are presumed to have been hunters in the EEA, would have been
served by a heightened ability to locate objects in space in order to target
animals the better to hit them with spear or rocks. Those men who were
better hunters would have been reproductively more successful,147 both
because they would have been preferred by choosy females (sexual
selection) and because they would have been more likely to survive (natural
selection) and reproduce for longer periods of time. Thus, selection
pressures would have led to the spread of the gene “for” spatial ability in
the male gene pool.148
In addition to cognitive differences, Richard Posner, Richard Epstein,
Kingsley Browne and others focus on data implying statistical differences
in preferences and other psychological tendencies between women and
men.149 Thus, for example, evolutionary theories of sexual selection predict
that men will differentially prefer certain traits in short- and long-term
potential mates, while females will differentially prefer other traits. In fact,
there are data which are consistent with these predictions. For example,
some studies show that men on average prefer mates who are young and
beautiful,150 while females on average prefer mates who are rich and high146. But see ANNE FAUSTO-STERLING, MYTHS OF GENDER: BIOLOGICAL THEORIES ABOUT
MEN AND WOMEN 26–36 (1992) (questioning the results of many of these studies, and arguing
that any difference in spatial ability is minuscule at best).
147. In evolutionary biology, “success” refers to reproductive fitness, which in turn refers
to both direct (offspring) and indirect (kin) reproduction. Thus, the most successful individuals
are those which produce the most offspring, which themselves go on to reproduce, as well as
help to ensure the success of close kin who share many of their genes. The sum total of this
direct and indirect success is referred to as “inclusive fitness.” E.g., Jones, Child Abuse, supra
note 4, at 1133–36 & nn. 43–45 (and sources cited therein).
148. This is a necessary simplification that tends to obscure some real questions about
evolutionary explanations of average differences. Because the only genetic difference between a
male and a female human being is that the male has a single Y chromosome that the female
lacks, any gene “for” a particular trait would either have to be located on the Y chromosome or
be expressed as a recessive sex-linked trait. This puzzle is usually explained with reference to
male and female sex hormones. See, e.g., SIMON BARON-COHEN, THE ESSENTIAL DIFFERENCE:
THE TRUTH ABOUT THE MALE AND FEMALE BRAIN 98–109 (2003) (discussing the effects of preand post-natal testosterone on the brain and on behavior).
149. See BROWNE, BIOLOGY AT WORK, supra note 4, at 53–67; POSNER, supra note 4, at 93;
Browne, Sex and Temperament, supra note 4, at 1016–50 (asserting that data supports the
existence of sex differences in the traits of risk-taking, status-seeking, aggressiveness,
nurturance, competitiveness, achievement-motivation, and dominance-assertion); Browne,
Women at War, supra note 4, at 80–88; Epstein, Nouns, supra note 145, at 990 (nurturing
instinct of women); id. at 992–93 (risk-taking preferences and competition).
150. The theory is that youth in females (but not as much in males) is a proxy for lifetime
reproductive potential and fertility, and that “beauty” consists of features that are correlated with
youth and health. See Buss, supra note 138, at 2 (“Features of physical appearance [of women]
associated with youth—such as smooth skin, good muscle tone, lustrous hair, and full lips—and
37:0321]
SEXUAL HARASSMENT LAW
371
status.151 And, especially relevant to employment law, men tend to be less
risk-averse, on average, and more competitive than are women.152 Finally,
according to evolutionary analyses, woman on average would be expected
to prefer to devote resources to childrearing and to domestic tasks, whereas
men’s average preferences would be weighted toward resource-gathering.
These average differences have in fact been found by some psychological
studies.153
Based upon these observed average differences, the legal analysis in this
genre generally proceeds as follows: (1) a certain empirical reality—for
example, occupational segregation by gender—has been observed and
documented; (2) this observed state of affairs can be explained, at least in
behavioral indicators of youth—such as high energy level and sprightly gait—have been
hypothesized to provide the strongest cues to female reproductive capacity”). It is often posited
that, for example, nice skin is a reasonable proxy for youth and good health (for example, lack
of infection and low parasite load), and that this is the evolutionary reason that men tend not to
find wrinkles or facial sores attractive in a mate. On the other hand, bad skin is rather more
often associated with reproductive age and raging hormones, and I have never heard an
evolutionary biologist to posit that acne is beautiful. For peer criticism and discussion of Buss’s
data and conclusions, see Open Peer Commentary, 12 BEHAV. & BRAIN SCI. 14–39 (1989)
(twenty-seven responses to Buss cross-cultural mate preference study). In addition, the use of
self-reported data in evolutionary science has been seriously challenged by biologists and
others. See Stephanie A. Shields & Pamela Steinke, Does Self-Report Make Sense as an
Investigative Method in Evolutionary Psychology?, in EVOLUTION, GENDER, AND RAPE 87
(Cheryl Brown Travis ed., 2003).
151. See Buss, supra note 138, at 12 (concluding that, in all of the cultures studied, females
rated the traits of ambitiousness, industriousness and high earning capacity in a mate more
highly than did males, and that for males, relative youth in a mate was more important than for
female selection of males); Fremling & Posner, supra note 4, at 1078 (“[M]en compete to
acquire resources, and success in that competition greatly influences their rank in the male
status hierarchy. Women compete for the high-status men, and the women who are successful in
this competition will have a high female status.”).
152. See BROWNE, BIOLOGY AT WORK, supra note 4, at 19–21. The average difference in
risk-aversion is said to partially explain the overrepresentation of men at both the highest and
lowest levels of social hierarchy. If employment (or political) success at the highest levels often
depends on a willingness to take risks, and if men as a group tend to be more willing to take
such risks, then one would expect more men than women to be corporate (or political) leaders.
Id. at 40–42. Likewise, the theory predicts that men are more likely than women to crash and
burn, both figuratively and literally. See id. at 20 (noting that young men, who show the greatest
risk-taking behavior, are disproportionately likely to die in traffic accidents).
153. See generally BUSS, supra note 59, at 19–72. These average differences have been
measured in studies that some critics describe as methodologically flawed. They also, however,
conform very closely to cultural gender ideals and myths. Asserted empirical differences, and
the cause of these differences, are of course analytically distinct issues. But see id. at 17–18
(arguing that there is an evolved human psychology that explains the universality of certain
gender stereotypes); STEPHEN E. RHOADS, TAKING SEX DIFFERENCES SERIOUSLY 21 (2004)
(describing studies by feminist researchers who “began their studies convinced that sex
differences were minimal and that societal forces caused those that existed,” but who changed
their views after examining the data).
372
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
part, as a result of the differential adaptive pressures that have acted to
shape the psychologies of males and females; (3) as a necessary corollary, it
is not best explained solely154 by other factors, such as discrimination or the
social construction of gender. To this point, the analysis remains
descriptive. The next step, however, is often normative: (4) because the
most parsimonious explanation for the particular state of affairs is evolved
average differences between men and women, society should refrain from
engineering an alternative result through legal rules.155
As an example, consider the explanation for the persistence of
occupational segregation by sex and the dearth of women in the highest
reaches of corporate leadership.156 Given the empirical reality of
occupational segregation by sex, how have some legal scholars brought
evolutionary analysis to bear? Both Kingsley Browne and Richard Epstein
have argued that such occupational segregation can in some significant
measure be explained by different average preferences and abilities in men
and women that are the result of sexual selection.157 Consequently, they
argue that neither illegitimate discrimination nor social construction of
gender necessarily or even best accounts for the data. Then, taking the
normative step, these scholars argue that legislation and policy designed to
correct these gender imbalances in the workforce might be misguided.158
154. The most thoughtful of these analyses do not take an either-or approach to biology and
environment; they recognize that the two are interactive. In critiquing the social constructivist
position, for example, Professor Browne argues not that culture is irrelevant, but rather that it
cannot explain everything. He offers instead “a biology that creates predispositions in
individuals that incline them more strongly in some directions than in others;” a biology on
which social and cultural forces then operate. See BROWNE, BIOLOGY AT WORK, supra note 4,
at 107.
155. See Browne, Biology, Equality, supra note 4, at 619–20 (arguing that, although
“recognition of sex differences does not compel an acceptance of all forms of sex
discrimination,” such recognition “does mean . . . that not all discrimination is invidious”).
Moreover, these arguments tend to stress the difficulty and the cost to human freedom of
attempting to alter “human nature.” See id. at 620 (“attempts to impose a sex-blind legal order
on a two-sexed species may not be successful without considerable coercion by the state” which
would “pose significant dangers to a free society”).
156. See GOOD FOR BUSINESS: MAKING FULL USE OF THE NATION’S HUMAN CAPITAL: FACT
FINDING REPORT OF THE FEDERAL GLASS CEILING COMMISSION (Mar. 16, 1995).
157. See Browne, Sex and Temperament, supra note 4, at 984 (arguing that the glass ceiling
is in large part “the product of basic biological sex differences in personality and temperament,”
and that “[t]hese differences have resulted from differential reproductive strategies that have
been adopted by the two sexes during human history”); Epstein, Nouns, supra note 145, at 992
(discussing sex differences in brain function caused by differential evolutionary payoffs).
158. See Browne, Biology, Equality, supra note 4, at 619 (stating that “[t]he existence of
sex differences in temperament suggests that the policy of attempting to achieve an
androgynous society through anti-discrimination laws, affirmative action programs, and
constitutional interpretation may be misguided.”). In other work, however, Professor Browne
37:0321]
SEXUAL HARASSMENT LAW
373
Here, as in much legal discussion which relies on evolutionary analysis
of sex difference, there has been a sharp focus on pinpointing and
quantifying distinctions between males and females at the expense of a
broader view of male and female behavior patterns. Yet it is precisely when
the more diffuse view is taken that evolutionary analysis can become an
extremely useful tool in understanding the conditions under which certain
behaviors are more likely to occur. Where, on the other hand, the focus is
not on context but rather on describing differences per se, there is a danger
that naturalistic assumptions will creep into the analysis. When an economic
perspective is combined with the evolutionary perspective, as is often the
case in legal analyses of biological sex difference,159 the naturalistic fallacy
is in danger of becoming more akin to a naturalistic imperative.
Others have noted the methodological and theoretical affinities between
evolutionary biology and economics.160 In particular, insofar as economic
theory holds that labor specialization and comparative advantage are
expected to lead to increased overall utility, sexual dimorphism might be
viewed as the ultimate efficient division of labor. Indeed, the phenomenon
of anisogamy (the difference between male and female gametes) is the
original division of labor from which all other sexual divisions ultimately
flow.161 The female and male sex cells are specialized to the functions of
nurturing and seeking, respectively.162 Once this specialization occurs,
has stressed that the policy decision regarding how society should treat sex difference is an issue
that is distinct from the questions whether those differences exist and, if they do exist, what the
cause(s) of those differences might be. E.g., BROWNE, BIOLOGY AT WORK, supra note 4, at 215.
159. Richard Posner and Richard Epstein, both noted law and economics scholars, are
emblematic of this approach.
160. See Baker, supra note 127 (discussing the affinities between economics and
evolutionary biology and highlighting especially the methodological similarities between the
two approaches).
161. In the usual account of how sexual reproduction between two sexes evolved and was
adaptive, in the beginning there was only one undifferentiated type of sexually reproducing cell.
Through random mutation and variation, some cells were larger and contained slightly more
nutrients, and some were smaller and faster but contained fewer nutrients. Once this
differentiation began, it would have tended to magnify so that eventually the sex cells evolved
into two types, one large and nutrient-rich, the other small and fast and containing only the
necessary genetic material. These two types are what we now refer to as the female sex cell
(egg) and the male sex cell (sperm): they are supremely specialized to their respective tasks. The
egg is specialized to nourishment; the sperm is specialized to locate eggs. See generally BOBBI
S. LOW, WHY SEX MATTERS: A DARWINIAN LOOK AT HUMAN BEHAVIOR 38–44 (2000)
(describing sex and strategies in reproduction).
162. See id. at 43. It should be noted that this rather uncontroversial observation about eggs
and sperm has often been extrapolated to describe women as passive and men as active, as well
as to create and reinforce separate spheres ideologies. Besides ignoring the relativity of these
role specializations, such assumptions also ignore a host of theoretical refinements and
observational data.
374
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
“subsequent changes in either gamete or carrier that enhance[] these
specific advantages will be favored by natural selection.”163 This role
specialization is fueled by natural selection because it confers a comparative
reproductive advantage upon those organisms which embody it: the
specialization is efficient, where efficiency is defined as the increased
ability to get a particular job done with the least investment. Those
organisms that fail to specialize (in relative terms) might be competent at
both nurturing and seeking, but will be less competent at either of those
tasks than those who have specialized. Pairs of nonspecialized organisms
would be expected to lose the sexual reproduction contest where they are
pitted against pairs of specialized nurturers and seekers.164
Economists as a rule prefer efficient outcomes; as a normative matter,
efficiency is the goal.165 Under economic reasoning, division of labor based
on role specialization is efficient. It follows, then, that such division of labor
would, all else being equal, be desirable. Unless it rests on coercion, such
specialization poses no normative problem. Role specialization, including
gender role specialization and labor role specialization, would be expected
to lead to greater total societal wealth and utility—to satisfy the conditions
of potential pareto optimality. And, if the roles166 are freely chosen, they
will lead to greater individual utility as well—they will satisfy the
conditions of actual pareto optimality.
This approach, while perhaps useful for examining certain kinds of
questions about efficiency, has several drawbacks. First, as noted, it tends to
lead to an assumption that what is evolved is efficient and therefore good, if
what has evolved is sex differentiation and role specialization. In addition,
this reasoning ignores more general patterns of male and female behavior as
163. Id. at 41.
164. For a graphic illustration of this dynamic as it plays out in the case of gamete
specialization, see id. at 40.
165. Economists define an outcome as efficient if it is pareto optimal, such that at least one
individual is better off and nobody is worse off. However, Pareto optimality is an elusive goal,
and many economists would agree that an outcome is desirable if it satisfies the conditions of
potential pareto optimality (or “Kaldor-Hicks” efficiency) “if the winners from the move benefit
more than the losers lose,” such that gainers could (though they generally do not) compensate
losers. See Howard E. Abrams, Economic Analysis and Unconstitutional Conditions: A Reply to
Professor Epstein, 27 SAN DIEGO L. REV. 359, 365 n.33 (1990). Though efficiency, in
economics, is a technical and not a normative concept, as a practical matter one would probably
be hard-pressed to find many economists who would not agree that the efficient outcome is
“good.” I thank Paul Rubin and Bill Carney for helpful comments on this point (though I do not
mean to suggest that either agrees with my conclusion).
166. I use labor in a broad sense to include women’s traditional labor inside the home
which is generally uncompensated. Both feminists and nonfeminist economists would, I believe,
concur in this usage.
37:0321]
SEXUAL HARASSMENT LAW
375
well as contextual variables that are crucial to a complete understanding of
both male and female preferences.
Thus far, law and evolution scholars who have analyzed Title VII sex
discrimination law have focused on the measurable differences between
men and women and have examined the issue from the perspective of
economic efficiency. However, the basic underlying assumptions of
evolutionary theory, and data regarding male manipulation-control behavior
toward females, have generally been overlooked or ignored. The larger
picture reveals that males might be expected to try to restrict female access
to economic resources as a method of constraining their mating choices.
Certainly one would expect the workplace to be a likely forum for this
behavior to be expressed. Preventing or restricting it would, arguably, serve
both libertarian free-market ideals and feminist notions of autonomy and
nonsubordination.
2.
Evolutionary Arguments and Hostile Work Environment Sexual
Harassment Law
This section considers the relevance of biological sex differences and
sexual selection theory to one particular manifestation of sex discrimination
in employment: that occurring when a person is subjected to a hostile
working environment based on sex. In this context, the average sex
differences in preferences and cognitive ability which might be relevant to
statistical proof of intentional discrimination or disparate impact are of no
moment, because hostile environment cases are not concerned with
assigning reasons for gender imbalances in the workplace. Instead,
evolutionary arguments about sex differences sometimes are relied on to
question the propriety of holding a defendant liable for behavior that, it is
argued, stems more from miscommunication167 than from blameworthy
intent. In an article applying evolutionary psychology to hostile work
167. See Browne, Seeking Roots, supra note 4, at 24–25 (stating that “there is substantial
evidence that aspects of a woman’s appearance may lead to miscommunication”; “there is much
room for misunderstanding in individual encounters”; “differences in perception that lead to
miscommunication are easily understood from an evolutionary perspective”; and “risk of
miscommunication is enhanced by the perception of many men that women often are just
‘playing hard to get’ and often mean ‘yes’ even if they say ‘no’”). Some feminists agree with
this description of the difference in the way that most men and most women perceive sexual
situations. See, e.g., Robin L. West, The Difference in Women’s Hedonic Lives: A
Phenomenological Critique of Feminist Legal Theory, 15 WIS. WOMEN’S L.J. 149 (2000)
(describing women’s lived experience of the fear of male sexual aggression and acquisitiveness,
and arguing that this difference in hedonic experience shapes a different reaction to sexual
encounters which men cannot understand).
376
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
environment sexual harassment law, Professor Kingsley Browne suggests
that “[m]uch of the conflict that is labeled ‘sexual harassment’ is traceable
in part to the fact that evolution has resulted in conflicting interests between
the sexes, which in turn have resulted in different sexual psychologies in
men and women.”168
These arguments frequently suggest a point that is the converse of that
made by courts and commentators about the issue of gender stereotyping.
Rather than regarding stereotypes as invidious or factually unsupported,
there is an implication that the stereotypes are accurate.169 Thus, with
respect to the potential miscommunication arising from a male harasser’s
perception that the victim is “playing hard to get,” Professor Browne notes
that “[a]lthough this notion is often referred to as a ‘myth,’ there is
168. Browne, Seeking Roots, supra note 4, at 8–9.
169. The Supreme Court has defined stereotype as a “frame of mind resulting from
irrational or uncritical analysis.” Nguyen v. I.N.S., 533 U.S. 53, 68 (2001). The dissent in
Nguyen criticized the majority’s definition, noting that the Court has recognized that “an
impermissible stereotype may enjoy empirical support and thus be in a sense ‘rational.’” Id. at
89 (Bryer, J., dissenting). For various formulations of the concept of stereotype, see Krieger,
supra note 96, at 1199 (relying on social cognition theory to define a stereotype as “a person’s
accumulated knowledge, beliefs, experiences (both direct and vicarious), and expectancies
regarding the schematized construct”); Michael S. Shin, Redressing Wounds: Finding a Legal
Framework to Remedy Racial Disparities in Medical Care, 90 CAL. L. REV. 2047, 2069 (2002)
(defining stereotypes as “beliefs about particular groups,” and stating that “implicit stereotypes
are subconscious mental representations of social categories—representations which involve
knowledge, beliefs, and expectations about social groups”) (citations omitted); Lis Wiehl,
“Sounding Black” in the Courtroom: Court-Sanctioned Racial Stereotyping, 18 HARV.
BLACKLETTER L.J. 185, 203 (2002) (defining stereotype as “a conformance to a group of
‘unvarying pattern . . . lacking any individuality’”) (quoting WEBSTER’S II NEW RIVERSIDE
UNIVERSITY DICTIONARY (1st ed. 1984)); see also Bem P. Allen, African Americans’ and
European Americans’ Mutual Attributions: Adjective Generation Technique (AGT)
Stereotyping, 26 J. APPLIED SOC. PSYCHOL. 884, 890 (1996); Jennifer L. Levi & Mary L.
Bonauto, Brief for Plaintiff-Appellant Lucas Rosa in the United States Court of Appeals for the
First Circuit Lucas Rosa v. Park West Bank and Trust Company on Appeal from the United
States District Court for the District of Massachusetts, 7 MICH. J. GENDER & L. 147, 155
(2001).
37:0321]
SEXUAL HARASSMENT LAW
377
substantial evidence that women do sometimes employ this tactic”170 and
that it is based on evolutionarily adaptive behavior patterns.171
It is apparent that many of the hypotheses of evolutionary psychology
conform to traditional social stereotypes regarding the proper roles of men
and women, the respective talents and abilities of men and women, and the
different psychologies of men and women.172 Indeed, critics often attack
evolutionary psychology on precisely this ground.173 In their view, the
congruence of theory with traditional social stereotypes reveals
discrimination behind the mask of science. On the other hand, proponents of
evolutionary psychology look at the convergence of the data, social
expectations, and their hypotheses and conclude that the theory is thus
supported. Social science explanations that discount the role of genetics and
biology are thus contrasted with “the common-sense intuition of the
170. Browne, Seeking Roots, supra note 4, at 25 n.118 (citing studies indicating that a
significant proportion of female college students have said that they did not want to engage in
sexual relations under circumstances in which they in fact were willing to do so); see also
PINKER, supra note 31, at 201–07 (discussing the concept of stereotyping, and arguing that
many stereotypes are based in fact and are a legitimate basis on which to make decisions absent
more individualized information); Browne, Women at War, supra note 4 (suggesting that
stereotypes about men and women could be a legitimate basis on which to bar all women from
military combat positions). For an explanation of stereotyping from a cognitive psychological
perspective, see Krieger, supra note 96.
171. Browne, Seeking Roots, supra note 4, at 25 n.118 (citing and quoting Linda Mealey,
Alternative Adaptive Models of Rape, 15 BEHAV. & BRAIN SCI. 397, 397 (1992) for the
proposition that “females are selected to be coy [so] that sometimes ‘no’ really does mean ‘try a
little harder’”).
172. Professor Browne states, in relation to his description of male-female differences in
temperament and ability:
[n]o doubt this sounds terribly sexist to some. The reader is asked to assume
that men are more competitive, more driven toward acquisition of status and
resources, and more inclined to take risks; women are more nurturant, risk
averse, less greedy, and less single-minded. These are familiar stereotypes to
us and to people around the world.
Browne, Sex and Temperament, supra note 4, at 981.
173. See, e.g., HRDY, supra note 26; Peggy Reeves Sanday, Rape-Free versus Rape-Prone:
How Culture Makes a Difference, in EVOLUTION, GENDER, AND RAPE 337, 342 (Cheryl Brown
Travis ed., 2003) (arguing that “the evolutionary argument provides scientific support for the
well-known popular belief in U.S. society that ‘boys will be boys’ and ‘girls ask for it’”);
MEREDITH F. SMALL, FEMALE CHOICES: SEXUAL BEHAVIOR OF FEMALE PRIMATES 10–11 (1993)
(criticizing mainstream sociobiology because it assumed, incorrectly in many cases, certain
roles of males and females, specifically, that men are promiscuous and aggressive and that
females are choosy and coy); Linda C. McClain, Symposium: Emerging Issues: Toward a
Formative Project of Securing Freedom and Equality, 85 CORNELL L. REV. 1221, 1230–31
(2000) (looking to evolutionary biology for enlightenment is problematic because “gender-role
expectations will color scientists’ ‘findings’ about the sexes”).
378
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
untutored that there is a fundamental difference between male and
female.”174
It is this intuitive sense on the part of many people that the stereotypes
are “right” in some general way that accounts for the appeal of the
arguments of evolutionary psychology and which make it all the more
important that its claims be subject to critical analysis.175 At the same time,
outright rejection of all biological explanation as either irrelevant or
incorrect risks winning some battles but losing the war.176 An engagement
with scientific theory and data offers the prospect of beginning a
conversation that might have much to offer women in the long run.177
3.
Summary
From the foregoing discussion of the application of evolutionary theories
to law, it is clear that only a narrow subset of the biological data offered by
the opponents of Title VII is specifically relevant to the issue of sexual
harassment in the workplace. Average statistical differences between men
and women in cognition, preference, and psychology, even if these do exist,
cannot support different treatment of individuals. Insofar as sexual
harassment sex discrimination based on the creation and maintenance of an
abusive working environment consists of disparate treatment
discrimination,178 average statistical differences between men and women in
174. Browne, Women at War, supra note 4, at 54 (“[T]here are some ideas so preposterous
that only an intellectual could believe them . . . .”).
175. Critiques of science have revealed, in many cases, the myriad ways in which the
preexisting cultural and social biases of the scientist can influence the questions that are asked,
the data that are observed or ignored, and the interpretation of those data. See generally articles
collected in FEMINISM AND EVOLUTIONARY BIOLOGY, supra note 30. However, these criticisms
need not, and should not, lead to a rejection of science. Rather, many scientists remain
committed to the scientific method and the ideals of objective fact and objective truth, but strive
to recognize and control their biases in designing their studies. Darwinian feminists Patricia
Adair Gowaty, Sarah Blaffer Hrdy, and Meredith Small are particularly notable in this regard.
See, e.g., Gowaty, supra note 58, at 917 (“Feminism made the experimental designs better.
Being self-conscious about my politics has helped to make my experiments better than they
might otherwise be, because I institute a variety of controls that others might also use, and
would no doubt use, if they were more aware of their own biases.”).
176. In addition, as Buss and others have pointed out, it begs the question of why such
patterns repeatedly occur across so many varied cultures.
177. Cf. Real Reform?, supra note 15, at 1982–83 (discussing the dilemma faced by “a
social reformer who realizes that the public threatened by her reform holds the power to prevent
that reform from occurring,” and suggesting an approach that would allow people honestly to
confront their biases).
178. Neither the disparate impact nor the disparate treatment theories of discrimination
apply very easily to the hostile work environment scenario. Because it is not necessary, in order
to prevail on a claim of hostile work environment discrimination, to prove any tangible job
37:0321]
SEXUAL HARASSMENT LAW
379
physical or mental ability and in preference patterns are irrelevant to the
treatment of individual plaintiffs.179 The only asserted average difference
that is potentially relevant to the hostile environment context is the
difference in the way that men and women initiate, perceive, and respond to
sexual overtures. Here, according to Professor Browne, a man might
reasonably believe that he is “flirting,” whereas a woman might reasonably
understand the same behavior as “abuse” or “harassment.”180 Even assuming
action or tangible psychological harm to the plaintiff, the McDonnell Douglas/Burdine disparate
treatment evidentiary framework does not comfortably apply to these claims. See Tex. Dep’t of
Commercial Affairs v. Burdine, 450 U.S. 248 (1981) (requiring, as part of plaintiff’s prima facie
case, proof of what amounts to a tangible employment action); McDonnell Douglas Corp. v.
Green, 411 U.S. 792, 802 (1973) (same). Unlike disparate treatment claims subject to the
McDonnell Douglas/Burdine burden-shifting formula, harassment claims typically involve
conduct that is admittedly illegitimate. The critical questions in these cases therefore are not
whether an otherwise lawful employment decision was made for an illegitimate reason, but
rather whether harmful behavior occurred “because of” the sex of the plaintiff, whether it was
unwelcome, and whether it rose to the required level of abusiveness and pervasiveness. See
Harris v. Forklift Sys., Inc., 510 U.S. 17 (1993); Meritor Sav. Bank v. Vinson, 477 U.S. 57
(1986). Likewise, hostile environment scenarios do not fit straightforwardly the disparate
impact framework, because they do not involve a neutral or otherwise legitimate employment
practice that only incidentally has disproportionate effects upon members of the protected class.
See, e.g., Watson v. Fort Worth Bank & Trust, 487 U.S. 977, 987 (1988) (“[T]he necessary
premise of the disparate impact approach is that some employment practices, adopted without a
deliberately discriminatory motive, may in operation be functionally equivalent to intentional
discrimination.”); Griggs v. Duke Power Co., 401 U.S. 424, 430 (1971) (holding that Title VII
prohibits “practices, procedures, or tests neutral on their face, and even neutral in terms of
intent, . . . [which] operate to ‘freeze’ the status quo of prior discriminatory employment
practices”). But see L. Camille Hébert, The Disparate Impact of Sexual Harassment: Does
Motive Matter?, 53 U. KAN. L. REV. (forthcoming 2004), available at
http://ssrn.com/abstract=555341 (last visited Feb. 16, 2005) (proposing a reconceptualization of
sexual harassment as disparate impact discrimination, in addition to the more common
understanding of sexual harassment as disparate treatment discrimination); Kearney, supra note
11, at 122 (arguing that “the phrase ‘facially neutral,’ though it appears as boilerplate in
hundreds of cases on the subject, does not appear anywhere in the statute” and should not be
read only to apply to “a policy or practice that an employer would acknowledge or even claim
pride in”); Kelly Cahill Timmons, Sexual Harassment and Disparate Impact: Should NonTargeted Workplace Sexual Conduct Be Actionable Under Title VII?, 81 NEB. L. REV. 1152,
1155 (2003) (arguing that nontargeted sexualized atmospheres are actionable only under a
disparate impact theory, and then “only if the conduct’s disproportionate impact on women is
great”).
179. These average differences are relevant, however, when statistical evidence is
permissible as proof of discrimination. The implications of the evolutionary argument in cases
utilizing statistical disparities as proof of discrimination are merely touched on here; a deeper
analysis of these questions is beyond the scope of this article, and will be addressed in a
forthcoming article.
180. Browne, Seeking Roots, supra note 4, at 24–28; see also BUSS, supra note 59, at 144–
48 (discussing studies demonstrating that men tend to infer sexual interest on the part of women
when it may not exist, and arguing that “[i]f over evolutionary history even a tiny fraction of
380
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
that such differences in perception exist, their cause is a matter of dispute.
Some would argue that such differences are the result of a sexist “double
standard” that penalizes female sexuality and encourages aggressive,
acquisitive male sexuality. Evolutionary psychologists, on the other hand,
explain these differences primarily as adaptations resulting from natural and
sexual selection.
The male-female differences advanced by evolutionary psychologists
might be relevant to hostile environment sexual harassment in this narrow
sense of measuring the moral blameworthiness of behavior that a man
arguably intends as flirtation and courtship but which a woman perceives as
threatening. However, as I argue in Part IV, the moral blameworthiness of
the individual harasser should not properly be considered from either a
policy or a doctrinal standpoint.181 Rather, the blameworthiness of the
employer, and not that of the individual harasser, logically should be the
focus of the liability inquiry in sexual harassment cases. Because persistent
“courting” behavior is often, and reasonably, perceived by some women as
threatening or abusive, the employer becomes blameworthy from a moral
standpoint once it has reason to know of the behavior and fails to prevent or
correct it.182 The argument, such as that advanced by Professor Browne, that
employers should not be held liable for “innocent” behaviors fails to
consider that the blameworthy act consists in failing to respond to a
foreseeable harm.
There is a final category of hostile environment cases in which the
asserted difference between men and women in the way that sexual
language, behavior, and innuendo is experienced might be relevant. That
category consists of those cases in which a sexually-charged environment is
alleged to be abusive, though not directed at any particular person and
indeed even if the identical environment existed before the person came to
work there.183 In such cases, the teachings of evolutionary psychology
these ‘misperceptions’ led to sex, then men would have evolved lower thresholds for inferring
women’s sexual interest”).
181. See infra text and notes 204, 206, 225.
182. In addition, this discrepancy in the way that males and females reasonably perceive
sexual situations highlights the importance of worker education about harassment: once
coworkers and supervisors learn that certain behaviors are often perceived by women as
threatening or abusive, they become blameworthy if they nonetheless continue to engage in
them.
183. See, e.g., Andrews v. City of Philadelphia, 895 F.2d 1469, 1485 (3d Cir. 1990);
Waltman v. Int’l Paper Co., 875 F.2d 468, 477 (5th Cir. 1989); Bennett v. Corroon & Black
Corp., 845 F.2d 104, 106 (5th Cir. 1988); Henson v. City of Dundee, 682 F.2d 897, 904 (11th
Cir. 1982); Robinson v. Jacksonville Shipyards, Inc., 760 F. Supp. 1486, 1522–1523 (M.D. Fla.
1991). But see Rabidue v. Osceola Ref. Co., 805 F.2d 611, 622 (6th Cir. 1986) (finding that a
sexual poster had a de minimis effect on the workplace, especially when considered in light of
37:0321]
SEXUAL HARASSMENT LAW
381
suggest that on average women would be differentially impacted by the
sexually-charged atmosphere of the workplace because of evolved sex
differences in psychology.184 This argument supports rather than
undermines the claim that the behavior constitutes harassment.185
For their part, legal feminists have often adhered to an equality model in
the realm of sex discrimination law.186 “Title VII was enacted, and was
originally litigated, under an equality-based account of discrimination . . . .
[under which] protected groups . . . were understood to be substantially
similar to the normative group—white men—for all purposes related to
employment.”187 The equality model is highly individualistic—it aims at the
injustice of treating an individual woman with a particular set of relevant
characteristics differently than a man with those same characteristics would
be treated.188 However, there are several areas in which Title VII doctrine
and theory have diverged from a strict equality model.189 In these areas,
what society condones and features “[in] newsstands, on prime-time television, at the movies
and other public [fora]”). For an extended discussion and analysis of this class of cases, see
Timmons, supra note 178, at 1207–35. I treat this category of cases supra in Part IV.A.
184. There is some support for the proposition that “bad language” (which includes words
with sexual connotations as well as words that describe bodily functions in a vulgar manner) is
more offensive, on average, to women than to men. See TIMOTHY JAY, CURSING IN AMERICA: A
PSYCHOLINGUISTIC STUDY OF DIRTY LANGUAGE IN THE COURTS, IN THE MOVIES, IN THE
SCHOOLYARDS AND ON THE STREETS 186–88 (1992).
185. That it supports or strengthens existing protections for women in the workplace is not
necessarily a reason to embrace this account of sex differences, and many would reject its
paternalism, reinforcement of the sexual double standard, and denial of women’s sexual agency.
See, e.g., Franke, supra note 6, at 746 (“Shutting down all sexual behavior seems like an
overreaction to the problem of sexual harassment, and requires some very disturbing
assumptions about the possibility of female sexual agency . . . .”).
186. For a discussion of the various strands of legal feminism, and a critique of a strict
adherence to the equality model, see FINEMAN, supra note 15, at 34–47.
187. Abrams, supra note 94, at 2518.
188. This individualistic focus has interesting parallels both in classical economics and in
mainstream evolutionary biology. The scientific biases that have resulted from a cultural
preoccupation with selfish individualism have been noted by biologists, see Lawton, et al.,
supra note 36, at 65–69, as well as by feminist legal theorists, see Baker, supra note 127, at
471–84. The symmetry between both of these areas of thought and the “rights” talk prevalent in
antidiscrimination jurisprudence is also striking. Over several areas of thought, there is a
theoretical tension between a focus on individuals and on groups; thus, in political theory,
communitarianism and liberalism; in discrimination theory, disparate treatment and disparate
impact; in evolutionary biology, gene selection and group selection.
189. For example, the treatment of pregnancy discrimination under Title VII recognizes the
relevance of perhaps the paradigmatic biological sex difference to claims of employment
discrimination. See 42 U.S.C. § 2000e(k) (2004) (defining the term “because of sex” in Title VII
to include “pregnancy, childbirth, or related medical conditions”). Note, however, that Title
VII’s treatment of pregnancy and childbirth generally tracks an equality model by analogizing
pregnancy to medical conditions faced by men and then comparing treatment by the employer.
In addition, sexual harassment discrimination has often been conceived, both in the case law
382
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
which include some forms of sexual harassment, one can discern as a
“minor theme”190 the notion that women and men are differently situated
with respect to the relevant treatment.
Even where these minor themes can be said to be operating, however,
feminist legal theorists often describe the relevant differences, or more
significantly the response to those differences, as socially constructed.191
The differences between a critical emphasis on the social creation of the
significance of sex on the one hand, and the evolutionists’ focus on the
adaptiveness of the social traits on the other, might seem a distinction more
of degree than of kind. Both groups, broadly speaking, agree that there are
certain basic biological differences between males and females. Both
groups, broadly speaking, view environmental factors as significant. It is
rather the way that biology and social constructs interact that is the subject
of disagreement.
First, for many critical scholars and feminists the subjective self is not an
unchanging, given entity that moves through an environment and is acted
upon by that environment. Rather, the social forces mold and change the
subjectivity itself.192 And, insofar as a person might be said to have some
and in the academic literature, as based in large part on sex differences, whether understood as
socially constructed or biological in nature. The highly influential dominance theory of sexual
harassment articulated by Catharine MacKinnon is based on a view that sexuality in the
workplace has very different implications for women and for men based on the social
construction of sexuality as a means of subordinating women. E.g., CATHARINE A.
MACKINNON, SEXUAL HARASSMENT OF WORKING WOMEN: A CASE OF SEX DISCRIMINATION
220–21 (1979).
190. This phrase is used by Professor Abrams. See Abrams, supra note 94, at 2517.
191. See id. at 2532 (advocating a social constructivist understanding of discrimination
whereby “biologically based qualities acquire their meaning through a process of social
construction,” both in the eyes of the subjects and the perpetrators of the discriminatory
conduct); cf. Tribe, supra note 42 (proposing a conception of law that, in the tradition of modern
physics, would account for the ways in which rules and behavior interact and influence one
another).
192. See, e.g., JUDITH BUTLER, EXCITABLE SPEECH: A POLITICS OF THE PERFORMATIVE 5
(1997) (“[I]t is by being interpellated within the terms of language that a certain social existence
of the body first becomes possible.”); Franke, supra note 6, at 693 (arguing that sexual
harassment is a performative disciplinary practice which “inscribes, enforces, and polices the
identities of both harasser and victim according to a system of gender norms that envisions
women as feminine, (hetero)sexual objects, and men as masculine, (hetero)sexual subjects.”).
As noted by Professor Abrams, this view offers an “unbiologized account”—it divorces gender
identity from biology as such. Kathryn Abrams, The New Jurisprudence of Sexual Harassment,
83 CORNELL L. REV. 1169, 1191–92 (1998). This point is more clearly apparent in the context
of race than of sex, insofar as the blurring of boundaries in the former is more frequent and
more visible. For example, when a person of mixed ancestry self-identifies as a particular race,
and is so identified culturally and/or legally, it seems clearer that a social choice is operating as
distinct from purely biological constructs. In the context of sex, the operation of these social and
political mechanisms is more hidden but not necessarily less real. See Katherine M. Franke, The
37:0321]
SEXUAL HARASSMENT LAW
383
fixed biological characteristics not molded by social forces, those
characteristics are imbued with social meanings that bear scant relation to
the biological givens. Social construction arguments thus emphasize the
primacy of cultural forces and highlight the unjust and inequitable ways that
they operate. In addition, these arguments frequently emphasize the hidden
power relationships that are created, enforced, and legitimized by the social
constructs that overlie biological givens.
In contrast, behavioral biology emphasizes the biological rootedness of
the social constructs.193 Indeed, the very term “sociobiology” implies just
that—the existence of biological and genetic forces that drive social
behavior by animals. Both social constructionists and evolutionists view
environment and culture as acting upon genes. It is the degree of flexibility,
and the normative evaluation of the consequences, that divide the two
approaches.
Putting aside, for the moment, this potentially irreconcilable difference
in perspective, the question remains: even accepting the premises of the
evolutionary argument for average sex differences in behavior and
psychology, what are the implications for the law of sexual harassment?
The above discussion reveals that, insofar as evolutionary explanations of
sex differences, as opposed to expected average male and female behavior
patterns, might be relevant to sexual harassment law, the nature of these
differences argue in support of current sexual harassment protections.
Alleged differences in the way that men and women perceive sexual
behavior suggest that such behavior does have a differential impact on
women in the workplace. In addition, the data and theory on differences
show that women reasonably experience persistent sexual advances as
harassment even where the male harasser might not view his behavior in
that way. To the extent that evolutionary psychology has something to say
about male-female difference, what it says is consistent with much current
thinking about the harms of sexual harassment.
Central Mistake of Sex Discrimination Law: The Disaggregation of Sex from Gender, 144 U.
PA. L. REV. 1, 26–31 (1995) [hereinafter Franke, Disaggregation] (discussing the analogy of
sexual identity and legal classification to racial identity and legal classification); cf. Jon Hanson
& David Yosifon, The Situation: An Introduction to the Situational Character, Critical Realism,
Power Economics, and Deep Capture, 152 U. PA. L. REV. 129, 165–66 (2003) (summarizing
social psychological literature showing that situation, rather than disposition, is the crucial
factor in determining human behavior).
193. See, e.g., David M. Buss, Author’s Response, 12 BEHAV. & BRAIN SCI. 39, 41 (1989)
(responding to peer critiques of cross-cultural mate preference study, and stating that “all
versions of the ‘culture hypothesis’ and ‘structural powerlessness’ hypothesis leave a
fundamental question unanswered: What is the origin of the economic inequality between males
and females that is found so pervasively across cultures? An evolution based model of
intrasexual competition provides one potential answer.”).
384
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
But evolutionary psychology speaks not only of sex differences; it has
much to say, also, about more generalized patterns of behavior and the
environmental contexts in which these would be likely to arise. These more
general patterns are relevant to the law of sexual harassment, and they have
been neglected by legal scholars employing evolutionary analysis to the
issue of sex discrimination. The balance of this Article focuses on these
more general patterns—the evolutionary forest, so to speak, rather than the
trees of male-female differences—and discusses the usefulness of
evolutionary explanations of these contextualized behavior patterns in the
analysis of sexual harassment doctrine.
IV. EVOLUTIONARY EXPLANATIONS FOR TYPICAL HARASSMENT BEHAVIORS
Sexual harassment takes a variety of forms. In order to analyze the
phenomenon, it is helpful to break it down into reasonably noticeable fact
patterns that may raise distinct questions from a biological and legal
perspective.194 Despite the fact that every abusive environment is in some
sense unique, and despite the courts’ extreme focus upon context in
evaluating abusiveness,195 it is possible to describe patterns of abuse and
harassment that are relevant to understanding the behaviors at issue and
194. Some legal scholars have argued that sexist harassment and sexual harassment should
not be distinguished for purposes of hostile work environment analysis. See Abrams, supra note
192, at 1217 (stating that, like Schultz, she views her theory of sexual harassment as broad
enough to include sexualized and nonsexualized conduct because both may be “a means of
preserving male control and entrenching male norms in the workplace”); Schultz, supra note 4,
at 1710–20. Whether because of a focus on harassment as a technology of sexism, see Abrams,
supra note 192, at 1217, or for more tactical reasons, see Schultz, supra note 4, at 1720–21
(arguing that courts’ tendency to separate the two types of harassment “weakens the plaintiff’s
case and distorts the law’s understanding of the hostile work environment by obscuring a full
view of the culture and conditions of the workplace”), these scholars have been suspicious of
approaches that deconstruct or de-aggregate harassment behaviors. However, one can recognize
that both kinds of harassment serve to subordinate women, and that both can contribute to a
pervasively abusive environment, while at the same time asking whether they might have
different proximate and ultimate causes and perhaps different prevention strategies. Cf.
Katharine K. Baker, Once a Rapist? Motivational Evidence and Relevancy in Rape Law, 110
HARV. L. REV. 563, at 569–89 (1997) (arguing that “rape” encompasses several different
behavioral patterns, and criticizing Federal Rule of Evidence 413 for its failure to account for
the various motivations behind different kinds of rape).
195. See, e.g., Harris v. Forklift Sys., Inc., 510 U.S. 17, 23 (1993) (“[W]hether an
environment is ‘hostile’ or ‘abusive’ can be determined only by looking at all the
circumstances.”). Justice Scalia, concurring in Harris, noted the unpredictable and highly
subjective nature of the Court’s contextual hostile environment standard, but stated that he
could not come up with a better test with which to judge whether particular conditions
constituted a hostile work environment in light of “the inherently vague statutory language.” Id.
at 24–25 (Scalia, J., concurring).
37:0321]
SEXUAL HARASSMENT LAW
385
evaluating these behaviors under the governing legal standard. This Part
describes these observable patterns of behavior, and then explains their
likely causes based on an evolutionary model. The next Part uses the
insights gained from these explanations to inform and re-evaluate the
applicable legal doctrine.
A. Recurring Patterns in Harassment Cases
Broadly speaking, the scenarios that constitute workplace sexual
harassment fall into three general categories. I label these scenarios,
respectively, “Give Me What I Want,” “Get Out of My Space,” and “Take It
Like A Man.” In the first type of case, a single harasser uses one or more of
a variety of tactics in an attempt to “get sex” from the victim.196 The facts of
Meritor Savings Bank, FSB v. Vinson197 fall into this category. This general
grouping would include quid pro quo cases198 as well as hostile environment
cases in which, for example, a coworker persistently propositions the
victim.199 Though there are myriad variations on the theme,200 the crucial
196. Some recent scholarship has sharply criticized courts’ reliance on a “sexual desire”
paradigm. See Cheryl L. Anderson, “Thinking Within the Box”: How Proof Models Are Used to
Limit the Scope of Sexual Harassment Law, 19 HOFSTRA LAB. & EMP. L.J. 125, 137 (2001)
(noting the “sexual desire” paradigm does little to discourage courts from viewing all malefemale harassment through this lens); Franke, supra note 6, at 714–24; Schwartz, supra note 11;
Julianne Scott, Student Scholarship, Pragmatism, Feminist Theory, and the Reconceptualization
of Sexual Harassment, 10 UCLA WOMEN’S L.J. 203, 220–22 (1999). These scholars point out
that much sexist harassment is not sexual, either in purpose or in method. In drawing the lines as
I do in this section, I do not mean to disagree with this proposition. It is of course true that much
harassment on the basis of sex is not accomplished through sexual conduct or words. In
addition, some scholars question whether even sexual behavior is necessarily aimed at sexual
relations, or rather is aimed at domination and power. See Schwartz, supra note 11, at 1763–66.
Again, I do not mean here to take a position on the issue of the relationship between sex, power,
subordination, and domination. I suggest only that, functionally, some harassing conduct is
aimed at sexual relations with the plaintiff, whatever the underlying motive or ultimate cause.
197. 477 U.S. 57 (1986). In Meritor, the plaintiff, Mechelle Vinson, alleged that her
supervisor, Sidney Taylor, “made repeated demands upon her for sexual favors” and that she
acquiesced because of a “fear of losing her job.” Id. at 60. Vinson testified at trial that
“Taylor . . . followed her into the women’s restroom when she went there alone, exposed
himself to her, and even forcibly raped her on several occasions.” Id.
198. I use the term quid pro quo here to include any situation in which the harasser uses an
implicit or explicit threat to influence the victim to accede to his sexual demands. This includes
those cases in which the victim acquiesces to the demand and thus can point to no tangible job
detriment. It also includes cases in which the victim refuses the demand but the threat is not
carried out. As discussed infra, this outline of the category of quid pro quo harassment differs
from the Supreme Court’s most recent formulation. I argue, however, that the Supreme Court’s
current definition of quid pro quo harassment is misguided. See infra at pp. 174–81.
199. See, e.g., Ellison v. Brady, 924 F.2d 872, 873–76 (9th Cir. 1991) (describing plaintiff’s
coworker who “pestered” her at work, “h[u]ng around her desk,” repeatedly asked her to lunch,
386
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
quality of cases in this category is that the harassing behavior implies the
intention or effect of coercing, extorting, or forcing sexual or other romantic
relations upon its target.201
In the second general pattern, one or more harassers “haze” one or a few
victims using sexual or nonsexual abuse, or some combination of sexual
and nonsexual abuse. Robinson v. Jacksonville Shipyards, Inc.202 is a typical
example of facts falling into this category.203 Here, the intent or effect of the
and wrote her disturbing notes and letters). The categories of quid pro quo and hostile working
environment sexual harassment were originally defined by Catharine MacKinnon. See
MACKINNON, supra note 189, at 32–47. They were explicitly adopted by the EEOC in 1980 in
its Guidelines on Sexual Harassment, and were implicitly adopted by the Supreme Court in
Meritor, 477 U.S. at 62. More recent cases, however, have limited the usefulness of the
distinction. See Burlington Indus., Inc. v. Ellerth, 524 U.S. 742, 751–53 (1998); see also
discussion, infra.
200. My approach here is analogous to evolutionary psychology’s understanding of cultural
variation. Evolutionary psychologists recognize variability in expression of human traits but see
at bottom some definable “human nature.” They stress that, though it is incredibly malleable
and variable, it is not infinitely so. See BROWNE, BIOLOGY AT WORK, supra note 4, at 1 (noting
that “despite the dizzying array of cultural practices chronicled by ethnographers, ranging from
the charming to the bizarre, we can be quite certain” that we will, for example, find no society
in which humans “live upside down.”); see also RIDLEY, supra note 3, at 3:
[A] psychiatrist can make all sorts of basic assumptions when a patient lies
down on the couch. He can assume that the patient knows what it means to
love, to envy, to trust, to think, to speak, to fear, to smile, to bargain, to
covet, to dream, to remember, to sing, to quarrel, to lie.
Id. In the same vein, though the particular expression of the behaviors is quite variable in the
hostile environment context, the basic patterns remain more or less constant and discernable.
201. I use the term “sexual relations” here in a broad sense, to include attempts to initiate
any kind of romantic relationship with the plaintiff. I recognize that use of the word “romantic”
in the context of a discussion of sexual harassment is problematic, but I am constrained by my
limited vocabulary and have not been able to come up with a better label to describe the desire
on the part of the putative harasser to initiate a nonplatonic relationship with the victim.
202. 760 F. Supp. 1486 (M.D. Fla. 1991). In Robinson, plaintiff Lois Robinson was one of
a handful of women working in a traditionally male shipbuilding shop. She and the few other
female employees testified that the workplace was suffused with sexually explicit “pinups” and
a generally sexualized atmosphere. After Robinson began complaining about this environment,
she was targeted for an array of sexualized and nonsexual abuse by several of her male
coworkers and supervisors. Id. at 1493–1502.
203. In Europe, this kind of harassment is referred to as “mobbing,” and there is much
academic and popular literature on the phenomenon. See Gabrielle S. Friedman & James Q.
Whitman, The European Transformation of Harassment Law: Discrimination Versus Dignity, 9
COLUM. J. EUR. L. 241, 247–63 (2003) (citing numerous sources in the European literature on
mobbing). Ironically, the term derives from animal behavior. “Mobbing” is used by
ornithologists to refer to “the joint assault on a predator too formidable to be handled by a single
individual in an attempt to disable it or at least drive it from the vicinity.” WILSON, supra note
32, at 46. European child psychologists and then industrial psychologists applied this concept of
animal behaviorism to humans. Friedman & Whitman, supra, at 247–48. Also called “moral
harassment,” mobbing is a form of “psychoterror” that is understood to include such behaviors
as “[r]efusing to communicate with an employee,” “[s]hunning . . . degrading working
37:0321]
SEXUAL HARASSMENT LAW
387
conduct is to humiliate or terrorize the target so as to induce him or her to
leave the workplace or to submit to the prevailing power structure. Though
the conduct is often sexual, the motivation is not primarily sexual.204 Often,
the hazing conduct takes the form of sabotaging and undermining the
victim’s ability to perform her job.205 As in Robinson, this type of
harassment is frequently initiated after the victim complains about other
forms of sexual material or behavior in the workplace.206
In the third category fall those cases in which one or more victims are
offended by words or conduct, often sexual in nature, that are not directed
specifically at the victim or victims. In Robinson, described above, the
workplace had been suffused with a “locker room” sensibility for years
prior to the entry of women into the traditionally all-male shipbuilding
positions. In another case, male firefighters brought a successful First
Amendment challenge to a city sexual harassment policy that prohibited
them from possessing and looking at pornographic materials in their private
spaces.207 Though more than one pattern may be present in a single case,
understanding the behavior patterns along these lines serves to make clearer
the ultimate causes208 behind the typical behavioral dynamics associated
with sexual harassment in the workplace.
conditions,” “[b]ullying, humiliations,” and “[s]landerous comments, insults, threats.” Id. at
246–47. For a comprehensive examination of the issue of workplace bullying from the
perspective of U.S. law, see generally David C. Yamada, The Phenomenon of “Workplace
Bullying” and the Need for Status-Blind Hostile Work Environment Protection, 88 GEO. L. J.
475 (2000).
204. Admittedly, creating these categories also creates some potentially thorny evidentiary
problems. How is a fact-finder to determine whether the harasser’s motivation is primarily
sexual, or rather is focused on power, humiliation, and exclusion? It seems, however, that the
factual issue is no more difficult—in fact is much less difficult—than the “because of sex”
inquiry currently required, and in most cases the motive can easily be inferred from the conduct.
Furthermore, both patterns imply legally-cognizable harassment and thus the categorization,
while helpful in order to apply the evolutionary models of behavior, is not crucial doctrinally
once the evolutionary analysis is completed.
205. Professor Vicki Schultz has collected and meticulously analyzed cases of this type.
She argues that, particularly in traditionally male workplaces in which the sex ratio is seriously
unbalanced, “a core element of [plaintiffs’] harassment is conduct having the aim or effect of
undermining [the women’s] work competence.” Schultz, supra note 4, at 1762. As Professor
Schultz demonstrates, this harassment takes many forms, but “sexual desire” is not its primary
motivation. See id. at 1759–61 .
206. Doctrinally, such conduct would give rise to a claim of retaliation. See generally 42
U.S.C. § 2000e-3(a) (making it an unlawful employment practice under Title VII to
discriminate against an employee “for making charges, testifying, assisting, or participating in”
any investigation or proceeding under Title VII); Armstrong v. Index Journal Co., 647 F.2d 441
(4th Cir. 1981); Shannon v. Bellsouth Telecommunications, Inc., 292 F.3d 712 (11th Cir. 2002).
207. See Johnson v. Los Angeles Fire Dep’t, 865 F. Supp. 1430, 1442 (C.D. Cal. 1994).
208. In evolutionary biology, ultimate cause is distinguished from proximate cause.
Ultimate cause refers to the evolutionary reason that a particular behavior or trait was adaptive;
388
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
In breaking down and analyzing harassing behaviors along these lines,209
I focus on the sexual or nonsexual nature of the functional motives of the
harasser. Thus, behavior directed at getting sex may sometimes be
obviously sexual in content, but it often is not. Likewise, behavior more
accurately understood as “hazing”210 or “mobbing” may employ sexuallycharged words or actions, and may even include sexual assault,211 but such
behavior is not necessarily, I will argue, primarily motivated by a desire to
get sex from the victim. However, as will become clear below, even in
those contexts in which the proximate aim of the harasser is to haze the
victim rather than to get sex from the victim, an evolutionary explanation
suggests that the ultimate cause of the conduct is “because of . . . sex.”212
B. Evolutionary Explanations
The first broad category of harassment cases, as outlined above, concerns
harassers whose conduct appears designed to procure sexual or romantic
relations from the victim. Among those whose aim is sexual in that sense,
the behaviors may range along a spectrum, from “flirting” to coercion to
rape. The other two broad categories are defined by their opposition to the
proximate causes, on the other hand, are the more immediate triggers that lead to the behavior
being expressed under particular circumstances. See Jones, Child Abuse, supra note 4, at 1128.
209. I use the terms “harassers,” “harassment,” and “victim” in this Part in order to describe
social, and not necessarily legal, conclusions. Cf. Heather Antecol & Deborah Cobb-Clark, The
Changing Nature of Employment-Related Sexual Harassment: Evidence from the U.S. Federal
Government, 1978–1994, 57 INDUS. & LAB. REL. REV. 443, 447 (2004) (“[F]ollowing standard
practice in this literature, our notion of sexual harassment is based on one or more experiences
of unwanted sexual behavior. It does not rely on individuals reporting themselves to have been
‘sexually harassed’ and does not necessarily fit with legal definitions.”); Browne, Seeking
Roots, supra note 4, at 8 (noting that the diversity of conduct that could constitute sexual
harassment makes statistical assertions about its prevalence misleading); Laura Beth Neilsen,
Situating Legal Consciousness: Experiences and Attitudes of Ordinary Citizens About Law and
Street Harassment, 34 L. & SOC’Y. REV. 1055 (2000).
210. By using the term “hazing,” I do not mean to minimize the seriousness of the behavior.
Hazing, across a range of social contexts, can result in serious injury or death and often
encompasses serious criminal offenses. See, e.g., Amanda Paulson, Hazing Case Highlights Girl
Violence, THE CHRISTIAN SCI. MONITOR, May 9, 2003, at 1 (describing hazing incident at a
suburban Chicago high school where girls were beaten and smeared with feces and fish guts);
Who Was Coach Protecting?, NEWSDAY, Dec. 21, 2003, at A08 (reporting violent hazing at a
boys’ football camp, including sodomy, beatings, and other “torture”).
211. See, e.g., Oncale v. Sundowner Offshore Serv., Inc., 523 U.S. 75, 77 (1998) (plaintiff
was threatened with rape, and was actually sexually assaulted by two of his harassers, in a
pattern that could be understood at least in part as “hazing” of the plaintiff).
212. See 42 U.S.C. § 2000e-2 (2004). As noted supra note 189, Title VII requires that
harassing behavior be “because of” the sex of the plaintiff in order to constitute sex
discrimination. Id.
37:0321]
SEXUAL HARASSMENT LAW
389
first: In the remaining examples, the harasser or harassers exhibit behaviors
that are not primarily or solely aimed at sexual relations with the victim.
Often, the victim appears to be targeted partly as a way of fostering bonds
within the rest of the group.213 Though the categories can, and often do,
overlap, it is useful to separate them in order to examine the relevant
behaviors more closely.214
In the latter categories, harassers exhibit behaviors that appear to
enhance their status within the group or to assert dominance over the victim
of the harassment. Many of the behaviors appear to be aimed at driving the
victim from the workplace, or communicating to the victim and others that
he or she is unwelcome.
1.
“Give Me What I Want”
An evolutionary understanding of that subset of behaviors that are
directly aimed at sexual relations with the victim of the harassment suggests
two significant conclusions. First, the evolutionary model reveals that words
and conduct aimed at establishing a sexual relationship are not only
“sexual” but also “because of” the sex of both the harasser and the victim of
harassment in a factual, but-for causative sense. Furthermore, an
evolutionary perspective focusing on ultimate causation renders both the
sexual orientation and the anatomical sex of the victim irrelevant to the
analysis. In other words, behavior that is directed at “getting sex” is, as a
descriptive matter, based on the sex of the plaintiff regardless of the actual,
proximate motive of the harasser. Finally, recent scholarship in evolutionary
biology and genetics on the role of constraint on female choice in sexual
selection has important implications for understanding these behaviors.
213. The experiences of victims in the military academies exemplify this type of abuse. See
Valorie K. Vojdik, Gender Outlaws: Challenging Masculinity in Traditionally Male
Institutions, 17 BERKELEY WOMEN’S L.J. 68, 68–75 (2002) (describing experiences of Shannon
Faulkner at The Citadel in terms that suggest that her victimization had more to do with
cementing bonds between the other cadets than with any overtly sexual agenda); see also
Browne, Women at War, supra note 4, at 97–102 (discussing the importance of male coalitions
in evolutionary theory, and arguing that women should be excluded from military combat
positions based on the importance of all-male groups) (claiming that it is necessary for men to
form groups); see generally LIONEL TIGER, MEN IN GROUPS (1969).
214. For example, a harasser might begin with the aim of having a sexual relationship with
the victim but then turn hostile after being rejected. Or, one or multiple harassers might sexually
threaten or assault a victim for the purpose of bonding with the group, forcing the victim to
leave, or demonstrating status and power. Some literature suggests that “gang rape” serves
largely a male bonding function. See generally PEGGY REEVES SANDAY, FRATERNITY GANG
RAPE: SEX, BROTHERHOOD, AND PRIVILEGE ON CAMPUS (1990); Claudia Card, Rape as a
Weapon of War, 11.3 HYPATIA 1 (1996).
390
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
As noted above, the theory of natural selection equates “fitness”
primarily with reproductive success.215 Reproductive success in males
depends primarily on access to fertile partners. In humans, as in other
mammals, the female of the species invests the greater differential
proportion of energy and time in offspring, and thus the female is regarded
as the limiting resource in the competition for mates. For these reasons,
evolutionary theory posits that, as a general matter, males will seek quantity
in sexual partners, while females will seek quality. Evolutionary scholars
typically describe males as tending to be “promiscuous,” and females as
tending toward “coy” or “choosy” behavior.
The model of the promiscuous male and the monogamous female216 has
been modified in recent years in order that it might be reconciled with
emerging observational and DNA evidence suggesting that females of many
species are not nearly so coy as previously believed.217 Similarly, there is
mounting data suggesting that males of some species, including humans,
have evolved under adaptive pressures for fidelity as well as for
promiscuity.218 Nonetheless, the traditional model remains dominant and has
been relied upon by legal scholars who argue that its teachings call the
prohibition against sexual harassment, along with other workplace
antidiscrimination doctrines, into question.219 This section therefore begins
215. See discussion supra at notes 55–58. The definition of fitness as individual
reproductive success is incomplete, however, because of the role of “inclusive fitness.” The
theory of inclusive fitness posits that the measure of an organism’s success encompasses the
success of close kin, since these relatives share significant percentages of genes. Thus, “kin
selection,” the reproductive success of parents, siblings, children, and other close relatives
affects the overall success of the genes of the individual. Kin selection, along with the theory of
“reciprocal altruism,” goes a long way to explaining the puzzle of altruism and cooperative
behavior in many species. See generally Robert L. Trivers, The Evolution of Reciprocal
Altruism, 46 QUART. REV. OF BIOL. 35 (1971) (reciprocal altruism); Hamilton, supra note 128
(kin selection); WRIGHT, supra note 111.
216. This view of sexual behavior is reflected in the rhyme penned by William James:
“Hoggamus higgamous, men are polygamous; higgamus hoggamus, women monogamous.”
ANGIER, supra note 141, at 355. James’ theory that human behavior was largely influenced by
instinct was eclipsed by the behavioral school of psychology for nearly a century. See RIDLEY,
supra note 2, at 316–20.
217. See HRDY, supra note 26; SMALL, supra note 173, at 3 (“[c]ontrary to what theory
suggests, female primates just keep on being highly sexual.”); RIDLEY, supra note 2.
218. See BUSS, supra note 59, at 43 (discussing the human universal of love and its relation
to female desire for mates who demonstrate commitment and fidelity); Charles C. Snowdon,
The “Nature” of Sex Differences: Myths of Male and Female, in FEMINISM AND EVOLUTIONARY
BIOLOGY, supra note 30, at 284–89 (describing the “high degree of [mate] fidelity by both
sexes” in studies of tamarins and the importance to offspring viability of paternal protection in
gorillas and other primates).
219. See BROWNE, BIOLOGY AT WORK, supra note 4, at 210–13; RICHARD A. EPSTEIN,
FORBIDDEN GROUNDS: THE CASE AGAINST EMPLOYMENT DISCRIMINATION LAWS 269–74
37:0321]
SEXUAL HARASSMENT LAW
391
by analyzing the behavior patterns involved in the first category—those
involving one harasser who seeks through his behavior to “get sex” from
the victim—by employing the standard evolutionary explanation of
differential parental investment and consequent male promiscuity and
female choosiness.220
Evolutionary theory understands much, if not most, male behavior as
ultimately directed toward reproduction.221 Thus, male conduct that is
explicitly aimed at asking, encouraging, coercing, or forcing romantic or
sexual relations upon a female is the most obvious instance of adaptive
behavior driven by Darwinian selection. Professor Kingsley Browne, in an
article arguing that hostile work environment harassment liability is
problematic because men and women understand the same behavior
differently based on differential selection pressures,222 states that contrary to
many feminists’ assertions that sex is used by men to gain power over
women,223 evolutionary understandings would posit the converse: Power is
used by men to gain sex.224
(1992); POSNER, supra note 4, at 90–92 (stating that men have stronger sex drives than women
for evolutionary reasons); Browne, Seeking Roots, supra note 4, at 12–28 (discussing
differences among the sexes); Fremling & Posner, supra note 4, at 1088–94 (relying on the
evolutionary model which posits male promiscuity and female choosiness).
220. As discussed infra at note 230 and accompanying text, some assumptions of this basic
model have been called into question. Here, however, I analyze the relevant behavior taking the
standard evolutionary model on its own terms and given the usual assumptions.
221. Evolutionary psychologists have sometimes been criticized by biologists for their
failure to recognize that much sexual behavior in animals might serve adaptive functions other
than reproduction. See Travis, supra note 28, at 6 (citing studies by primatologists Frans de
Waal and Frans Lanting that demonstrate that sexuality in bonobos, one of humans’ two closest
genetic primate relatives, serves largely a social and group cohesiveness function that goes far
beyond its simple reproductive function). See generally DE WAAL & LANTING, supra note 62
(describing the mating habits of the bonobo apes and their similarities to humans).
222. Browne, Seeking Roots, supra note 4, at 8–9. Professor Browne argues that, because of
men’s and women’s different minds, it is morally questionable to hold men liable for behavior
which they reasonably view as courting, whereas females may reasonably view the identical
behavior as “harassment.” Id. at 37–39. He thus understands much sexual harassment as in fact
consisting of simple miscommunication resulting from male/female differences.
223. This idea is generally associated with radical feminism. See, e.g., ANDREA DWORKIN,
INTERCOURSE (1988) (equating heterosexual intercourse with rape); MACKINNON, FEMINISM
UNMODIFIED, supra note 99, at 3 (arguing that, in our society, violence against women is
eroticized so that sex becomes violence and violence becomes sex). However, certain strands of
the theory have become widespread in the popular culture with, for example, the general
acceptance of the notion that rape is a crime primarily of violence rather than of sex. See SUSAN
BROWNMILLER, AGAINST OUR WILL: MEN, WOMEN AND RAPE 376–77 (1975); Baker, supra
note 26, at 238 (“If the law is to curb the behavior of rapists, it must recognize that rape is used
by men as a conscious, instrumental tool to gain access to power, prestige, and community.”);
Georgia Network to End Sexual Assault, Basic Facts About Rape and Sexual Assault, at
http://www.gnesa.org/sexual_assault/facts.html (last visited Mar. 2, 2005) (“Rape is motivated
392
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
Because it is so clear under evolutionary explanatory models that
behavior of men toward women that is directed at sexual relations is highly
adaptive, and that it is therefore both ultimately and proximately tied to the
sex of the object of the behavior, these cases tend to be unproblematic from
a causation standpoint.225 The more difficult cases in the first category have
arisen under circumstances of same-sex harassment (and in the elusive but
no longer purely theoretical case of the bisexual, or “equal opportunity,”
primarily out of . . . a need to feel powerful by controlling, dominating, or humiliating the
victim.”). In this view, rape is an act by which individual men exert power over individual
women, and by which men as a group use fear of rape to subordinate women as a group. Rape
as a War Crime, at www.vday.org/contents/violence/glossary/rapeaswarcrime (last visited Mar.
2, 2005). Recent evolutionary scholarship has challenged this conception of rape as violence
divorced from sex. See THORNHILL & PALMER, supra note 26; Jones, Rape, supra note 4, at
897–98; Randy Thornhill and Nancy W. Thornhill, The Evolutionary Psychology of Men’s
Coercive Sexuality, 15 BEHAV. & BRAIN SCI. 363 (1992); see also Browne, Seeking Roots,
supra note 4, at 47–48 n.221 (citing and quoting sources in the “power versus sex” debate in the
rape literature).
224. See Browne, Seeking Roots, supra note 4, at 48 (“Throughout human history, men
have used power as a way of obtaining sex . . . .”).
225. The theory of workplace sexual harassment as Title VII sex discrimination is generally
understood to have grown out of the work of radical feminists, most notably Catharine
MacKinnon. See Meritor Sav. Bank v. Vinson, 477 U.S. 57, 63 (1986) (using same general
framework for sexual harassment as Title VII discrimination described in MacKinnon’s work);
MACKINNON, supra note 189, at 57–59; see also Schultz, supra note 87, at 2078–82 (2003)
(recounting history and development of the “sexualized understanding of harassment”).
Professor MacKinnon’s theory of sexual harassment as sex discrimination was grounded in her
notion that much male-female sexual behavior is a method by which men subordinate women.
Because sexual harassment doctrine was born of this dominance and subordination model of
male-female sexuality, its growing pains have often been manifested in cases that do not fit
easily into this model. Thus, cases in which abuse is sexist but nonsexual and cases in which the
harassment is not by a male or males against a female have presented the most difficult
theoretical and doctrinal issues. Feminist scholars writing in the 1990s addressed these issues in
several articles reconceptualizing sexual harassment law. These reconceptualizations have
moved away from or refocused MacKinnon’s antisubordination model such that nonsexual
harassment, same-sex harassment, and harassment based on nonconformity to prevailing gender
norms, come within the sexual harassment prohibition. See Abrams, supra note 192, at 1172;
Franke, Disaggregation, supra note 192, at 95 (“Title VII should recognize the primacy of
gender norms as the root of both sexual identity and sex discrimination, and thereby the law
should prohibit all forms of normative gender stereotyping regardless of the biological sex of
any of the parties involved.”); Franke, supra note 6, at 693 (arguing that sexual harassment is
sex discrimination because it is a “technology of sexism” that enforces feminine gender norms
upon women and masculine gender norms upon men); Schultz, supra note 4, 1692–96 (rejecting
the “sexual desire-dominance paradigm” and arguing that it should be replaced with a
“competence-centered paradigm”). One potential difficulty with a theory that anchors
harassment in gender norm enforcement is that, because gender norms are so pervasive, liability
is potentially limitless. It is unclear how one might distinguish between legitimate and
illegitimate gender norms. Thus, the theory would sweep all workplace harassment within the
concept of sexual harassment.
37:0321]
SEXUAL HARASSMENT LAW
393
harasser).226 More particularly, in the wake of Oncale v. Sundowner
Offshore Services, Inc.227 courts have had to confront explicitly the issue of
discriminatory causation in those cases in which the behavior appears
clearly to be sexual (and even directed toward “getting sex”), yet the sexual
orientation of the harasser is unclear.228 Thus, it is useful to examine the
way in which evolutionary theory might understand and explain the
behaviors underlying those cases.
Thus far I have assumed that in cases in which a man propositions, flirts
with, coerces, or sexually assaults a woman, it is obvious that such behavior
is aimed at “getting sex.” Courts have similarly understood such factual
226. See Hamm v. Weyauwega Milk Prod., Inc., 332 F.3d 1058, 1068 (7th Cir. 2003)
(“‘[s]ex stereotyping’ should not be regarded as a form of sexual discrimination, though it will
sometimes . . . be evidence of sex discrimination.”) (Posner, J., concurring); Lack v. Wal-Mart
Stores, Inc., 240 F.3d 255, 261–62 (4th Cir. 2001) (finding judgment for defendant where
harasser used vulgar language with both female and male harassees; court held that the
harassment was therefore not based on sex); Nichols v. Azteca Rest. Enters., Inc. 256 F.3d 864,
874–75 (9th Cir. 2001) (holding that a man had a claim for sexual harassment because of his
effeminate nature and that the Price Waterhouse rule, protecting women discriminated against
for not being sufficiently feminine, also applies to man who is not sufficiently masculine);
Holman v. Indiana, 211 F.3d 399, 401 (7th Cir. 2000) (finding “equal opportunity harasser” not
covered by Title VII where supervisor sexually harassed and propositioned both a husband and
wife); Bianchi v. City of Philadelphia, 183 F. Supp. 2d 726, 735–37 (E.D. Pa. 2002) (denying a
firefighter’s claim of sexual harassment where he could not prove that the harassment was due
to gender stereotypes and not sexual orientation); Oiler v. Winn Dixie La., Inc., 89 Fair Empl.
Prac. Cas. (BNA) 1832 (E.D. La. 2002) (rejecting claim of discrimination by man who crossdressed when not at work; court found firing was not based on gender stereotypes but on his
cross-dressing, which is not protected by Title VII).
227. 523 U.S. 75 (1998).
228. As when, for example, a male coworker or supervisor “gooses” other men, or in cases
involving the notorious “bisexual harasser.” See EEOC v. Harbert-Yeargin, Inc., 266 F.3d 498,
503 (6th Cir. 2001) (hearing male-male harassment suit in which “the practice of ‘goosing’
occurred among male employees . . . on a daily basis”); Holman, 211 F.3d at 403, 405 (holding
that “Title VII does not cover the ‘equal opportunity’ or ‘bisexual’ harasser, [then], because
such a person is not discriminating on the basis of sex” when married couple in same workplace
claimed that male supervisor sexually harassed them both). Some of the most interesting and
original scholarship on sexual harassment appears partly driven by the intuition that same-sex or
sexually-indiscriminate harassment cases should not be left in the dust as the Title VII caravan
disappears over the horizon. Though Congress has repeatedly rejected sexual orientation
discrimination per se as a valid claim under Title VII, some courts have held that sex role
enforcement harassment constitutes sex discrimination and not sexual orientation discrimination
and thus may state a proper cause of action under Title VII. See Butler v. Ysleta Indep. Sch.
Dist., 161 F.3d 263, 270 (5th Cir. 1998) (citing Professor Franke for the proposition that sexual
harassment must be a practice that is founded in the “service of hetero-patriarchal norms”); Doe
v. City of Belleville, 119 F.3d 563, 578 (7th Cir. 1997) (citing Professor Franke, and finding
that sexual stereotypes drove the harassment of two male brothers), vacated by 523 U.S. 1101
(1998); Polly v. Houston Lighting & Power Co., 825 F. Supp. 135, 137 (S.D. Tex. 1993).
394
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
scenarios in this way,229 and as noted above evolutionary theory also
understands the behavior as ultimately aimed at sexual access to
reproductive females.230 Where, on the other hand, the subject and object of
the behavior are not male and female, respectively, some courts have had
difficulty with the issue whether the behavior, though sexual, was “because
of” sex. Justice Scalia’s statements in Oncale are illustrative of this
thinking. In Oncale, Justice Scalia, writing for the Court, suggested that in
most cases, in order to satisfy Title VII’s requirement that the harassment be
“because of” the sex of the victim, a plaintiff in a same-sex harassment case
would be required to demonstrate that the harasser was homosexual.231
An evolutionary approach, however, would explain male harassment of
another male, done by means of sexual threats, coercion, assault, or
innuendo, as based on the sex of both the harasser and the victim regardless
of the other sexual behavior of the harasser. The same conclusion results
regardless of the sexual orientation of the victim.232 Evolutionary models
229. Cf. Schwartz, supra note 11, at 1699 (discussing prevalence of per se causation finding
in cases in which behavior was sexual in nature). For criticism of the “desire-based ‘but for’
causation theory” on the ground, among others, that it imports heterosexual assumptions of
behavior, see id. at 1719–21, and sources cited therein.
230. Professor Browne likewise understands the behavior in this way. See Browne, Seeking
Roots, supra note 4, at 9 (noting that “[a]n evolutionary perspective suggests [that] . . . men use
power instrumentally to obtain sex”).
231. Oncale, 523 U.S. at 79–80. The Court suggested as examples two other evidentiary
avenues of proof of the causation element. First, a plaintiff could offer evidence that a person of
the same sex directed nonsexual harassment towards the plaintiff on the basis of gender
animosity. Id. at 80 (“A trier of fact might reasonably find such discrimination, for example, if a
female victim is harassed in such sex-specific and derogatory terms by another woman as to
make it clear that the harasser is motivated by general hostility to the presence of women in the
workplace.”). Second, a plaintiff could offer evidence demonstrating that, in a mixed-sex
workplace, one sex was treated differently than the other. Id. at 80–81 (“A same-sex harassment
plaintiff may also, of course, offer direct com-parative [sic] evidence about how the alleged
harasser treated members of both sexes in a mixed-sex workplace.”). Professor Schwartz has
argued that the thrust of the Court’s opinion “invites the federal courts to embark on some
potentially very ugly lines of factual inquiry” regarding whether a man who sexually harasses
another man is “‘really’ a homosexual.” Schwartz, supra note 11, at 1745. Along with the
wisdom of such an approach, the relevance of the inquiry is also doubtful. “Most men who
sexually assault other men identify themselves as heterosexual.” For Men Only: For Male
Survivors of Sexual Assault, at http://www.utexas.edu/student/cmhc/booklets/maleassault/
menassault.html (last visited Mar. 2, 2005); see also Stop Prisoner Rape, The Basics on Rape
Behind Bars, at http://www.spr.org (last visited Feb. 19, 2005) (“[A] typical male prison rapist
chooses a victim on the basis of ‘the weakness and inability of the victim to defend himself.’”).
232. Questions about the cause, nature, and definition of homosexuality are certainly
beyond the scope of this article. What I attempt to do here is merely to outline the way an
evolutionary biologist might explain homosexuality, and to argue that such an explanation, by
its own terms, suggests that the conduct at issue in male-male harassment cases satisfies the
“because of . . . sex” requirement of Title VII. There is much recent scholarly work regarding
the nature and causes of homosexuality. See, e.g., DEAN HAMER & PETER COPELAND, THE
37:0321]
SEXUAL HARASSMENT LAW
395
also support the view that the behavior of the “bisexual harasser” is
ultimately based upon the sex of the harassed individual even though the
harasser targets both males and females. These understandings follow from
a description of this type of male behavior as driven ultimately by the drive
to foster alliances with other males so as ultimately to increase reproductive
success.233
Homosexuality has presented an obvious puzzle for evolutionists.
Various theories have been advanced to explain what appears on its face to
be a maladaptive behavior.234 One hypothesis centers on the notions of
inclusive fitness235 and kin selection. This theory posits that homosexuality
evolved by kin selection, because by forgoing direct reproduction and
SCIENCE OF DESIRE: THE SEARCH FOR THE GAY GENE AND THE BIOLOGY OF BEHAVIOR 17–38
(1994) (presenting evidence in favor of the existence of a “gay gene”); JIM MCKNIGHT,
STRAIGHT SCIENCE?: HOMOSEXUALITY, EVOLUTION AND ADAPTATION 8–14 (1997); R.C.
Kirkpatrick, The Evolution of Human Homosexual Behavior, 41 CURRENT ANTHROPOLOGY 385
(2000) 389–90 (citing numerous works); see also BUSS, supra note 59, at 250–56 (describing
homosexuality as one of the remaining “mysteries of human mating” and summarizing current
adaptationist theories); POSNER, supra note 4, at 98–108; RIDLEY, supra note 3, at 263–65, 279–
80 (describing theory of Oxford scientists Laurence Hurst and David Haig that “gay gene”
might be in the mitochondrial DNA inherited from the mother, and might be adaptive for
females because “[i]t effectively sterilizes males, causing the diversion of inherited wealth to
female relatives”).
233. Historically, the most powerful males left a disproportionate number of offspring. See
RIDLEY, supra note 2, at 198–202 (describing the research of Laura Betzig and noting that
“these harems [maintained by male rulers throughout most of recorded history] could hardly
have been more carefully designed as breeding machines, dedicated to the spread of emperors’
genes”). Men, along with the males of many species, form political alliances in order to gain
power. “From Hannibal to Bill Clinton, men gain power by putting together coalitions of
allies. . . . The rewards, for other animals, are largely sexual. For men?” RIDLEY, supra note 2,
at 197. Part and parcel of “putting together coalitions” is excluding certain individuals—an
alliance is meaningless except as it exists in opposition to something else; cf. FRANS DE WAAL,
PEACEMAKING AMONG PRIMATES 10–23 (Harv. Univ. Press ed., 1989) (arguing that the concept
of cooperation is meaningless without the concept of conflict).
234. The prevalence, therefore, of homosexual behavior calls into question the whole
enterprise of evolutionary psychology. Within biology, it supports Gould’s broader evolutionary
approach that sees many specific behaviors not as adaptations but as byproducts of other, more
general adaptations. E.g., Gould & Lewontin, supra note 1. Incidentally, my use of the word
“maladaptive” is in no way intended to convey any moral implication. Rather, it is a
sociobiological term that describes behavior that does not further, or positively hinders,
reproductive fitness. Individuals who engage exclusively in same-sex relations, like celibate
individuals, do not directly pass on their genes to future generations. It is only in this respect
that the behavior is nonadaptive or maladaptive. For a concise discussion of this issue from an
evolutionary and moral perspective, see WRIGHT, supra note 111, at 384–86.
235. Inclusive fitness is the sum total of an organism’s direct reproductive fitness and its
indirect fitness, which encompasses any increase in the reproductive success of the individual’s
kin attributable to the efforts of the individual, discounted by their degree of consanguinity.
E.g., Jones, Child Abuse, supra note 4, at 1133–36 nn. 43–45, and sources cited therein.
396
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
focusing energy on caring for and providing resources to close relatives, an
individual’s overall genetic fitness could be increased. A related theory
suggests that parental manipulation might have played a part in the
evolution of homosexual behavior, such that certain familial lineages could
increase their reproductive success by encouraging homosexual behavior in
sons with lower reproductive potential.236 Neither of these theories appears
to have much empirical support.237
A third hypothesis is that homosexual behavior evolved as a particular
application of the phenomenon of reciprocal altruism.238 In general, the
theory of reciprocal altruism holds that seemingly altruistic behavior (that
is, behavior that appears to benefit others at the expense of the individual),
can be explained by virtue of the selfish gene model of evolution by
examining the larger systematic benefits to individuals of helping others
who then return the favor.239 Coalition-building and male-male alliances are
one way for individuals to benefit through reciprocal altruism. Recent
animal studies demonstrate that “coalitions of males are found in a number
of species,” including turkeys, lions, woodpeckers, and dolphins.240 In these
nonhuman animals, male alliances are quite clearly tied to reproductive
success. According to this explanation, then, male homosexuality might
have evolved as a byproduct to male alliance-building traits.
Based upon the evolutionary models employed by scholars such as
Posner, Epstein, and Browne, male behavior toward other males that is
sexual in nature might be understood in much the same way as the kind of
male-male behavior that the Oncale Court characterized as “horseplay.”241
236. Note that most of these theories focus on the evolution of male homosexuality. There
has been much less study of female homosexuality. See Kirkpatrick, supra note 232, at 388; see
also Evelyn Blackwook, 41 CURRENT ANTHROPOLOGY 398, 398 (2000) (commenting on
Kirkpatrick article).
237. See Kirkpatrick, supra note 232, at 391–92.
238. Reciprocal altruism is the mechanism by which most evolutionary theorists explain
seemingly altruistic behavior not otherwise explainable by kin selection. See generally
DAWKINS, supra note 36; Robert Trivers, The Evolution of Reciprocal Altruism, 46 Q. REV.
BIOLOGY 35, 35–39 (1971); WRIGHT, supra note 111.
239. See generally Trivers, supra note 238; Robert Axelrod & W.D. Hamilton, The
Evolution of Cooperation, 211 SCIENCE 1390 (1981).
240. RIDLEY, supra note 2, at 195–96; see FRANS DE WAAL, CHIMPANZEE POLITICS: POWER
AND SEX AMONG APES 119–23 (1982) (describing male alliances among chimpanzees in a
captive group).
241. See Oncale v. Sundowner Offshore Serv., Inc., 523 U.S. 75 (1998):
[a] professional football player’s working environment is not severely or
pervasively abusive, for example, if the coach smacks him on the buttocks as
he heads onto the field––even if the same behavior would reasonably be
experienced as abusive by the coach’s secretary (male or female) back at the
office.
37:0321]
SEXUAL HARASSMENT LAW
397
This scenario therefore falls more easily within the second pattern of
behavior identified in this section—that directed ultimately at creating or
fostering male alliances. As understood by current evolutionary theory,
male sexual behavior toward other males can properly be understood as
ultimately adaptive based on its role in reciprocal altruism and coalitionbuilding. Because the evolutionary model explains the conduct by virtue of
the sex of the object of the behavior, it supports the proposition that malemale sexual behavior is closely tied to the sex of the plaintiff even where
the harasser considers himself to be heterosexual or engages in heterosexual
behavior in other contexts.242
2.
“Get Out of My Space”
The second relevant behavior pattern in harassment cases involves
conduct that appears to have the purpose or effect of intimidating,
humiliating, or otherwise “hazing” the victim so as to make that person
either leave the workplace or submit to the demands (whether explicit or
implicit) of the dominant group. Actors enforce a particular workplace
culture, and use various means to do so. As previously noted, such behavior
is often sexual or suggestive because sexual or suggestive conduct tends to
be an effective tool of humiliation or intimidation.243 Nonsexual tactics,
however, are often used as well. These might be obviously sexist, but they
need not be.244
Id. at 81.
242. Sociological and historical data suggest that many persons engage in homosexual
behavior while at the same time defining themselves as heterosexual. See Halley, supra note 32,
at 525 (noting that various studies show “that men affirm their identities as heterosexual even
when they acknowledge having recent same-sex contacts”); Kirkpatrick, supra note 232, at
388–92 (discussing studies).
243. In addition to often being perceived as humiliating or degrading by the victim,
sexually explicit words or conduct are frequently perceived as physically threatening. See
Elizabeth Schoenfelt et al., Reasonable Person Versus Reasonable Woman: Does it Matter?, 10
AMER. UNIV. J. GENDER SOC. POL’Y & L. 633, 648 (2002).
244. For example, part of the conduct that contributes to the hostile environment might
consist of statements such as “women belong at home” or “this is a man’s job.” See, e.g.,
Deborah Epstein, Can a “Dumb-Ass Woman” Achieve Equality in the Workplace? Running the
Gauntlet of Hostile Environment Harassing Speech, 84 GEO. L.J. 399, 415 (1996) (collecting
cases). Or the tactics might be more general and not inherently sexist, but nonetheless directed
to forcing the victim out of the workplace. See, e.g., O’Rourke v. City of Providence, 235 F.3d
713, 729–30 (1st Cir. 2001) (ruling that “incidents of nonsexual conduct -- such as work
sabotage, exclusion, denial of support, and humiliation -- can in context contribute to a hostile
work environment . . . .”); Williams v. General Motors Corp., 187 F.3d 553, 561 (6th Cir. 1999)
(holding that nonsexual pranks, such as gluing the plaintiff’s toolbox to her desk, allowed as
evidence of anti-female animus); King v. Bd. of Regents of Univ. of Wis. Sys., 898 F.2d 533,
398
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
Based upon an evolutionary model, such hazing or exclusionary behavior
can be understood as driven, at least in part,245 by the sex of the victim of
the behavior. The proximate aim246 can generally be understood as either
getting the victim to leave, or getting the victim to conform and cooperate.
In either event, evolutionary psychology would explain the behavior as
ultimately driven by sexual selection and, therefore, as “because of” the sex
of the victim.
Humans are members of a highly social species for which group
cooperation is and has been critical for survival.247 Whether one adheres to a
group-selection or a gene-selection model of natural selection, evolutionary
theories of selection describe social coalition and bonding behaviors as
highly adaptive.248 Primate infants, in particular, require a prolonged period
of care for survival. Writings in evolutionary psychology typically
understand much sexually-based role and labor divisions as stemming from
535 (7th Cir. 1990) (noting plaintiff’s harassment claim included allegations of unequal
treatement such as a heavier workload, lower salary, repeatedly negative evaluations, rudeness,
and interference with her tenure process); see also Schultz, supra note 4, at 1764, stating:
Sometimes [undermining a woman’s competence at work] takes the form of
deliberate sabotage of a woman’s work performance, such as stealing a
policewoman’s case files, informing a lab worker that faulty equipment is
sound, falsifying medical records to make it appear as though a female
surgery resident made an error, or simply assigning her tasks that are
impossible to accomplish. (citations omitted).
Id.
245. In disparate treatment cases in which “mixed motives” are alleged, the plaintiff may
prevail if she demonstrates that sex was “a motivating factor for any employment practice, even
though other factors also motivated the practice.” 42 U.S.C. § 2000e–2(m) (2000). If the
employer then proves that the same employment action would have been taken in the absence of
the impermissible motivation, the plaintiff is limited to declaratory and injunctive relief and
attorney’s fees. See 42 U.S.C. § 2000e–5(g)(2)(B).
246. In evolutionary biology, “proximate” cause is distinguished from “ultimate” cause. For
definitions of these terms as used in biology, see supra note 208.
247. Indeed, enforced social isolation is considered among the most severe punishments,
and has been shown often to lead to psychosis and suicide among those subjected to it. See
Mary Beth Pfeiffer, A Death in the Box, N.Y. TIMES, Oct. 31, 2004, Section 6 (Magazine), at 48,
52 (describing the “debate over the effects of isolation on even a normal human psyche” in the
context of a discussion of the use of solitary confinement to control mentally ill prison inmates).
248. Banishment from the community was a historic punishment for certain crimes, and
was considered essentially equivalent to death. See, e.g., Stogner v. California, 539 U.S. 607,
624–25 (2003) (noting the English tradition of using banishment as punishment for treason);
Smith v. Doe, 538 U.S. 84, 98 (2003) (contrasting punishments such as public shaming and
banishment to the mere dissemination of public information on sex offenders pursuant to
“Megan’s Laws”); Vance v. Terrazas, 444 U.S. 252, 271 (1980) (finding that a man should not
lose his citizenship, and characterizing banishment as “a fate universally decried by civilized
people.”) (Marshall, J., concurring in part and dissenting in part) (quoting Trop v. Dulles, 78 S.
Ct. 590, 599 (1958)).
37:0321]
SEXUAL HARASSMENT LAW
399
this dependency of the infant on the mother.249 In most species in which the
infant requires some significant post-natal parental investment in order to
survive, parental investment theory predicts that the mother will tend to
specialize in this task for several reasons, all stemming from the fact that
she internally gestates the offspring.250 First, in these species, the mother but
not the father can be certain that she is investing in offspring that carry her
genetic material. For the father, there is always some element of uncertainty
about paternity. This fact, combined with the mother’s greater initial
parental investment in the form of the nutrients contained in the egg and her
internal gestation, create a situation in which she has a relatively greater
incentive to invest further in the offspring.
Along with the natural dependency of the human infant or child, there
also exists a high level of interdependence among adults in human
groups.251 In the EEA, males are thought to have hunted cooperatively.
Males are also imagined to have warred cooperatively, both attacking and
defending in socially-bonded alliances.252 Those individuals exhibiting
stronger social skills could be expected to have increased their reproductive
fitness. Female grouping, too, can be explained as an adaptive behavior.253
249. Cf. FINEMAN, supra note 15, at 72.
250. In mammals, the mother also nurses the offspring once born; for nonhuman animals as
well as humans throughout all but a fraction of human history, the baby would die if the mother
did not make this further parental investment.
251. This is an obvious and fundamental observation, but its moral implications have
sometimes been overlooked. Evolutionary biology has often been criticized for its
individualistic bias; the tenets of nineteenth century social Darwinism cast a long shadow. Yet
even if human social interdependence is explained by virtue of “gene selfishness,” the result can
be a humanistic, inclusive, and optimistic view of human nature. See WRIGHT, supra note 111.
One view posits human moral progress in the progressive expansion of the “moral circles” by
which people define the beings subject to worth and empathy. See PETER SINGER, THE
EXPANDING CIRCLE: ETHICS AND SOCIOBIOLOGY (1981) (discussed in PINKER, supra note 31, at
166–67). In fact, Darwin himself seems to have understood morality in this way. He wrote:
As man advances in civilisation, and small tribes are united into larger
communities, the simplest reason would tell each individual that he ought to
extend his social instincts and sympathies to all the members of the same
nation, though personally unknown to him. This point being once reached,
there is only an artificial barrier to prevent his sympathies extending to the
men of all nations and races.
WRIGHT, supra note 111, at 340 (quoting Charles Darwin, circa 1882).
252. See Browne, Women at War, supra note 4, at 98–101 (citing numerous sources).
253. Though not as extensively treated as male group behavior, e.g., TIGER, supra note 213;
Browne, Women at War, supra note 4, at 97, female group behavior has been viewed as
adaptive in several respects. Female primates and other mammals may group together and thus
protect themselves and their infants from male violence. See Sarah L. Mesnick, Sexual
Alliances: Evidence and Evolutionary Implications, in FEMINISM AND EVOLUTIONARY BIOLOGY,
supra note 30, at 207, 216 (noting that “[f]emale coalitions are most likely to evolve in species
that reside in groups and are long-lived so that benefits can be reciprocated among coalition
400
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
The existence of social norms, including gender norms, is a human
universal.254 “By the time they are 3 years old, boys and girls around the
world participate in different activities, show different behavioral styles,
play more with same-sex peers, and avoid opposite-sex peers.”255 Though
the contents and details of social norms show enormous cross-cultural
variation—indeed, this variation is at the heart of what we mean by the very
term “culture”—that such norms are more or less strongly enforced by
formal or informal social practices is not seriously questioned.256 A
significant subset of social norms consists of gender norms—cultural
stereotypes concerning the appropriate behavior, appearance, and separate
roles (including labor roles) of men and women.257
The issue of the enforcement of social norms, including gender norms,
has often been a locus of disagreement between social scientists and others
who adhere to a social constructivist view of gender, on the one hand, and
evolutionary psychologists who view much stereotypically gendered
behavior as flowing from evolutionary adaptation, on the other. One
interesting finding of the child development psychological literature is that
young boys and girls who engaged in “cross-gender” activities were treated
members.”) (citing numerous sources); Barbara Smuts, Male Aggression Against Women: An
Evolutionary Perspective, 3 HUMAN NATURE 1, 16–18 (1992). In addition, females may help
one another with childcare and food-gathering. See, e.g., HRDY, supra note 30, at 282. One
theory explains female menopause as an adaptation that allowed older women to help younger
female relatives with childrearing tasks, thus furthering the grandmothers’ own reproductive
fitness. See Kristen Hawkes et al., Hazda Women’s Time Allocation, Offspring Provisioning,
and the Evolution of Long Postmenopausal Life Spans, 38 CURRENT ANTHROPOLOGY 551, 560–
61 (1997); Kristen Hawkes et al., Grandmothering, Menopause, and the Evolution of Human
Life Histories, 95 PROC. NAT’L ACAD. SCI. 1336, 1336–39 (1998); see also HRDY, supra note
30, at 273–83 (discussing “the grandmother hypothesis,” which she suggests should more
appropriately be called “the prudent mother hypothesis” because it encompasses fitness effects
upon a mother’s own offspring in addition to those of her children).
254. See DONALD E. BROWN, HUMAN UNIVERSALS (1991) list of universals reprinted in
PINKER, supra note 31, at app. 437 (“male and female . . . seen as having different natures”).
255. Beverly I. Fagot et al., Theories of Gender Socialization, in THE DEVELOPMENTAL
SOCIAL PSYCHOLOGY OF GENDER 65, 79 (Thomas Eckes & Hanns M. Trautner eds., 2000).
256. See Brant Wenegrat et al., Social Norm Compliance as a Signaling System: Studies of
Fitness-Related Attributions Consequent on Everyday Norm Violations, 17 ETHOLOGY AND
SOCIOBIOLOGY 403, 404 (1996) (noting that “[w]hile norms themselves and the cultural
emphasis placed on compliance may vary, similar processes of compliance with social norms
have been observed in all societies.”).
257. See Linda B. Epstein, Note, What is a Gender Norm and Why Should We Care?
Implementing A New Theory in Sexual Harassment Law, 51 STAN. L. REV. 161, 176 (1998)
(defining gender norms as “descriptive or normative gender stereotypes”); Eagly et al., supra
note 104, at 123; Franke, supra note 6, at 693 (explaining sexual harassment as “a disciplinary
practice that inscribes, enforces, and polices the identities of both harasser and victim according
to a system of gender norms that envisions women as feminine, (hetero)sexual objects, and men
as masculine, (hetero)sexual subjects”).
37:0321]
SEXUAL HARASSMENT LAW
401
dissimilarly, with the boys experiencing much more negative
consequences.258 Girls who played with boys or with boy-preferred toys
were not given increased negative feedback by other girls or by boys;
however, boys who played with girls or with girls’ toys were taunted and
rejected by other boys even after modifying their cross-gendered behavior.
“The type of negative feedback these boys received was of a different
quality from that received by other boys and included gender derogatory
terms such as sissy boy, baby boy, and so forth, which were almost never
applied to other children.”259 Furthermore, the boys’ behavior, unlike the
girls’, was responsive only to feedback from other boys; the boys “ignored
both the positive and negative feedback from girls and teachers.”260
If these patterns are widespread, and if this sex difference in peer group
gender norm enforcement persists over time and across cultures, then one
might wonder whether it represents an evolved psychological trait that
would have been adaptive in the EEA. In other words, one might ask
whether enforcement of gender norms, and in particular male peer group
enforcement of masculine gender norms against other males, is an
evolutionary adaptation of the human male psychology. It might indeed
have been highly adaptive for males in particular, and under particular
circumstances, to police other males in their social group for adherence to
stereotypically masculine behaviors and to exclude from their social groups
males who did not conform to these (now) stereotypical traits. Males who
were not sufficiently aggressive, or strong, or brave, would have been a
liability to the other males in the context of group hunting or warring.
Viewed in this light, gender norm enforcement by males against other males
would satisfy any “but for” causation requirement in the statute.261
Therefore, if, as behavioral biologists claim, males have strong
evolutionary tendencies to group together and to exclude females and “outgroup males” from these groups, then male behavior toward a female victim
in this second category of cases is easily explained by evolutionary theory
as “because of” the sex of the victim. Where the victim is male, the
argument is more complex but leads to the same conclusion though by an
alternate route. Where males in a group act to coerce or exclude other
258. See Fagot et al., supra note 255, at 65; Beverly I. Fagot, Consequences of Moderate
Cross-Gender Behavior in Preschool Children, 48 CHILD DEVELOPMENT 902, 902 (1977).
259. Fagot et al., supra note 255, at 80.
260. Id. at 81, 85; see also Beverly I. Fagot, Beyond the Reinforcement Principle: Another
Step Toward Understanding Sex Role Development, 21 DEV. PSYCHOL. 1097 (1985).
261. This analysis would be consistent with the approach of those scholars who have
argued that harassment based on gender norm noncompliance by “effeminate” men is sex
discrimination under Title VII. See Case, supra note 6, at 47–51; Franke, supra note 6, at 693.
402
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
males, they do so for evolutionary reasons that are tied to the sex of the
victim.262
3.
“Take It Like a Man”
In the final category of cases, a plaintiff complains of hostile
environment harassment where the words or conduct that she finds abusive
are not directed specifically at her. Indeed, in some cases the words or
conduct that the victim finds abusive were common currency in the
workplace before she entered.263 Under these circumstances, evolutionary
sexual selection models would support the argument that such an
environment would on average be more offensive and threatening to women
than to men.264
262. Biologists who study, for example, primate behavior consistently understand the
actions of the animals that they observe as closely based on the sex of the other individuals with
which those animals are interacting. In other words, the aggression of a male towards a female
is viewed differently from the aggression of the same male towards another male. See, e.g.,
FRANS DE WAAL, supra note 240, at 88. In fact, an animal behaviorist would likely view as
nonsensical the attempt to describe interactions between conspecifics without considering the
sexes of the animals involved.
263. See, e.g., Ocheltree v. Scollon Prods., Inc., 308 F.3d 351, 374–75 (4th Cir. 2002)
(finding defendant entitled to judgment as a matter of law where plaintiff entered workplace
permeated with crude sexual language and behavior, but such atmosphere preexisted her entry
and evidence did not demonstrate that the atmosphere changed in any way following her entry),
vacated by 335 F.3d 325 (4th Cir. 2003) (en banc); Rodgers v. Western-Southern Life Ins. Co.,
12 F.3d 668, 674 (7th Cir. 1993) (stating that courts may consider the background “lexicon of
obscenity” that existed before plaintiff entered the workplace).
264. An evolutionary analysis is thus consistent with the position that the abusiveness of the
environment should be determined by reference to a “reasonable woman” rather than a
“reasonable person” or “reasonable man” standard. See Browne, Seeking Roots, supra note 4, at
28 (“For very fundamental reasons, the sexes have differing views of sexuality. Where a man
might see ‘opportunity,’ a woman sees ‘danger.’”). Empirical data supports the view that words
and behavior that are sexual in nature or that describe body parts or functions in a vulgar way
are more offensive, on average, to women than to men. Of course it is entirely plausible to
maintain that this average difference is a result of social learning. It is also plausible to explain
the difference as a result of sexual selection pressures. My argument here does not purport to
answer this question of causation, but only to note that on its own terms the evolutionary thesis
would have to accept the proposition that women as women may be differentially impacted by a
sexualized or vulgar atmosphere. Professor Browne, though generally critical of the hostile
environment sexual harassment cause of action, has stated that “[i]f a biological perspective can
contribute anything to sexual harassment policy, it must be the insight that a ‘reasonable person’
standard is meaningless. At least when it comes to matters of sex and sexuality, there are no
‘reasonable persons,’ only ‘reasonable men’ and ‘reasonable women.’” Id. at 31. On the other
hand, he asserts that a “focus on the plaintiff’s perspective systematically privileges the
woman’s view, even if the two parties are equally responsible for miscommunication.” Id. at
38–39. The appropriate standard of reasonableness in sexual harassment cases has been the
topic of much scholarly and judicial discussion. See, e.g., Robert S. Adler & Ellen R. Peirce,
37:0321]
SEXUAL HARASSMENT LAW
403
It is in this class of cases that the evolutionary perspective is most useful
as a tool not for answering the question of causation but for highlighting the
uncertainty in the doctrine of causation in discrimination law.265 Absent a
clear articulation of what causation means in the context of sex
discrimination in general and hostile environment discrimination in
particular, it is impossible to know whether conduct, assumed to be abusive
to the victim, constitutes unlawful discrimination. In the prior two
categories of cases, it was sufficient to pose the question whether the
harasser, for whatever reason, treated the victim badly because of the
victim’s sex. In those cases, evolutionary theory suggested that the behavior
of the individual harasser was directed at the victim based at least in part on
her or his sex. Here, by assumption, there is no behavior that is directed at
the victim266 and thus the question must necessarily be posed differently.267
The Legal, Ethical, and Social Implications of the “Reasonable Woman” Standard in Sexual
Harassment Cases, 61 FORDHAM L. REV. 773, 825–27 (1993) (questioning whether it is fair to
men to hold them to a standard to which they cannot relate); Leslie M. Kerns, A Feminist
Perspective: Why Feminists Should Give the Reasonable Woman Standard Another Chance, 10
COLUM. J. GENDER & L. 195, 196–97 (2001) (making a feminist argument in support of using
the reasonable woman, as opposed to the reasonable person, standard); Elizabeth L. Shoenfelt et
al., Reasonable Person Versus Reasonable Woman: Does It Matter? 10 AM. U. J. GENDER SOC.
POL’Y & L. 633, 641–58 (2002) (arguing that the difference between the “reasonable woman
standard” and the “reasonable person standard” is merely semantic, but that social scientists are
right to consider gender differences when determining reasonableness and that other changes,
such as educating male jurors on the perspectives of women, would be more fruitful than
arguing over the standard itself); E. Gary Spitko, He Said, He Said: Same-Sex Sexual
Harassment Under Title VII and the “Reasonable Heterosexist” Standard, 18 BERKELEY J.
EMP. & LAB. L. 56, 83 (1997) (arguing that the need for the “reasonable woman standard” is less
now that Title VII litigants are entitled to a jury trial, thus ensuring that some of the triers of fact
will be female, as opposed to a federal judge, who is more likely to be male).
265. The meaning of the intent or causation element in anti-discrimination law has been the
subject of much academic inquiry. See Krieger, supra note 96, at 1176; Evan Tsen Lee &
Ashutosh Bhagwat, The McCleskey Puzzle: Remedying Prosecutorial Discrimination Against
Black Victims in Capital Sentencing, 1998 SUP. CT. REV. 145, 150–60; Ann C. McGinley, !Viva
La Evolución!: Recognizing Unconscious Motive in Title VII, 9 CORNELL J.L. & PUB. POL’Y
415, 448 (2000); Schwartz, supra note 11, at 1709, 1714–17; Michael Selmi, Proving
Intentional Discrimination: The Reality of Supreme Court Rhetoric, 86 GEO. L. J. 279, 288
(1997); David A. Strauss, Discriminatory Intent and the Taming of Brown, 56 U. CHI. L. REV.
935, 937–38 (1989); D. Don Welch, Removing Discriminatory Barriers: Basing Disparate
Treatment Analysis on Motive Rather Than Intent, 60 S. CAL. L. REV. 733, 740 (1987); Rebecca
Hanner White & Linda Hamilton Krieger, Whose Motive Matters?: Discrimination in MultiActor Employment Decision Making, 61 LA. L. REV. 495, 527 (2001); Michael J. Zimmer, The
Emerging Uniform Structure of Disparate Treatment Discirmination Litigation, 30 GA. L. REV.
563, 564 (1996).
266. The term “directed at” is susceptible to more than one interpretation. I use the term
here to imply motive or intention on the part of the actor. If he hangs up a girlie calendar in the
workplace in order to enjoy the photographs, it is not “directed at” the plaintiff (nor at anyone,
for that matter, except perhaps himself). If he hangs it up in order to bother the plaintiff, then it
404
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
In the nondirected cases, the motivations (whether conscious,
unconscious, “evolutionary,” proximate, or otherwise) of the various actors
whose behavior directly contributes to the abusive environment are most
obviously irrelevant to the issue whether the environment is discriminatory.
Their behavior is not discriminatory “treatment” in any intelligible sense of
the term.268 Consequently, this species of case points to the overall logic of
is no longer an undirected act. Similarly, if sexist jokes are inadvertently overheard by the
plaintiff, the words are not “directed at” her. On the other hand, if the jokes are told loudly in
her presence so that she will overhear them, they are directed at her. Compare Sanchez v. City
of Miami Beach, 720 F. Supp. 974, 977 n.9 (S.D. Fla. 1989) (holding that pornographic pictures
posted in police station were “directed” in that they singled out female employees), and
Robinson v. Jacksonville Shipyards, Inc., 760 F. Supp. 1486, 1522–23 (M.D. Fla. 1991) (same),
and Stair v. LeHigh Valley Carpenters Local Union No. 600, 813 F. Supp. 1112, 1114–16 (E.D.
Pa. 1993) (calendars of naked women evinced “implicit” intent to discriminate on the basis of
sex), with Johnson v. County of L.A. Fire Dep’t, 865 F. Supp. 1430, 1438 (C.D. Cal. 1994)
(finding that the presence of Playboy magazines was unrelated to the creation of a sexually
hostile working environment), and Rabidue v. Osceola Ref. Co., 805 F.2d 611, 622 (6th Cir.
1986), cert. denied, 481 U.S. 1041 (1987) (finding that posters and pin-ups of nude women “had
a de minimis effect on the plaintiff’s work environment”).
267. The question must be posed differently, that is, unless one concludes that this class of
cases by definition does not implicate Title VII. Some courts have so held, at least in those cases
in which the abusive environment preexisted plaintiff’s entry therein. See Gross v. Burggraf
Constr. Co., 53 F.3d 1531, 1538 (10th Cir. 1995) (stating that the evalution of sexual
harassment claims must take into account “environment[s] where crude language is commonly
used by male and female employees” and that “[s]peech that might be offensive or unacceptable
in a prep school faculty meeting, or on the floor of Congress, is tolerated in other work
environments.”); Rabidue, 805 F.2d at 622 (finding, notoriously, that a claim premised in part
upon preexisting vulgarity and pornography in the workplace must be “considered in the context
of a society that condones” such displays); see also Smith v. Northwest Fin. Acceptance, Inc.,
129 F.3d 1408, 1414–15 (10th Cir. 1997) (holding that the existing workplace culture is relevant
to a determination of harassment, but finding that this factor cut in favor of the plaintiff because
of the demure and professional workplace culture that preexisted her arrival). Most courts,
however, have repudiated the Rabidue holding and have reasoned that defendants “cannot use
their discriminatory pasts to shield them from the present-day mandate of Title VII.” Smith v.
Sheahan, 189 F.3d 529, 535 (7th Cir. 1999); accord Williams v. Gen. Motors Corp., 187 F.3d
553, 564 (6th Cir. 1999); Robinson, 760 F. Supp. at 1524–25. Several legal scholars have
criticized judicial treatment of the “prevailing workplace” issue. See, e.g., Jane L. Dolkart,
Hostile Environment Harassment: Equality, Objectivity, and the Shaping of Legal Standards, 43
EMORY L.J. 151, 199 (1994); Susan Estrich, Sex at Work, 43 STAN. L. REV. 813, 846 (1991).
Courts holding that the preexisting environment cannot give rise to a harassment claim rest this
result on the argument that there can have been no intention to treat the plaintiff differently
where the treatment was uniform and did not change as a result of the plaintiff’s arrival in the
workplace. This reasoning applies with equal force, however, to cases in which the conduct is
not directed specifically at the plaintiff though it arises after plaintiff’s arrival in the workplace.
In either event, the focal issue should not be the intention of the actor toward the plaintiff
personally, but rather the differential effect of the behavior on the plaintiff by virtue of her sex.
268. There is one caveat to this argument. It may very well be the case that persons in
certain workplaces cultivate a “macho” or sexualized environment for the conscious or
unconscious purpose of preserving that workplace as a sphere of male exclusivity. In those
37:0321]
SEXUAL HARASSMENT LAW
405
re-directing the intent or causation inquiry to where it more reasonably
belongs: to the question of the employer’s intent as revealed by the acts or
omissions of itself or its agents in preventing or responding to the abusive
conduct.269
4.
The Significance of the Workplace Context Under Evolutionary
Theory
The fact that harassing behavior has been situated in or around the
workplace270 takes on heightened significance when an evolutionary
instances, one might imagine that the environment is “directed” at the (eventual) victims
notwithstanding the fact that it is in place before they arrive to complain. Cf. Rogers v. EEOC,
454 F.2d 234, 239 (5th Cir. 1971) (recognizing in the first case explicitly to countenance a cause
of action under Title VII for hostile environment harassment based on racial hostility, the
“distinct possibility that an employer’s [seemingly nondirected] discrimination may constitute a
subtle scheme designed to create a working environment imbued with discrimination and
directed ultimately at minority group employees”).
269. See infra text and note 333 and accompanying text.
270. In some cases, part of the conduct that comprises a hostile work environment occurs
outside the physical workplace. See, e.g., Dowd v. United Steelworkers of Am., 253 F.3d 1093,
1101 (8th Cir. 2001) (describing racially harassing conduct occurring outside plant in
connection with strike picket line); Moring v. Ark. Dep’t of Corr., 243 F.3d 452, 454–55 (8th
Cir. 2001) (detailing offensive sexual conduct by supervisor toward plaintiff after hours and on
a business trip); Ellison v. Brady, 924 F.2d 872, 874–75 (9th Cir. 1991) (finding coemployee
sent plaintiff letters, called her, and came to her home); Aldridge v. Kansas, No. 96-2382-JWL,
1997 WL 614323, at *9–10 (D. Kan. Sept. 10, 1997) (unpublished decision) (finding that
actionable conduct included lewd and inappropriate remarks and physical pursuit of plaintiff
that occurred outside the workplace); McGuinn-Rowe v. Foster’s Daily Democrat, No. 9462350, 1997 WL 669965, at *4 (D. N.H. July 10, 1997) (unpublished decision) (holding that
harassment by supervisor that took place in a local bar could be part of the pattern of conduct
that creates a hostile working environment). Some courts distinguish between supervisory and
coworker conduct for purposes of deciding whether harassment that occurs outside the
workplace may properly be considered. See Durham Life Ins. Co. v. Evans, 166 F.3d 139, 153
n.9 (3d Cir. 1999) (finding harassment outside workplace actionable where plaintiff was forced
to work with the harasser outside of the workplace); Turner v. Reynolds Ford, Inc., 145 F.3d
1346 (10th Cir. 1998) (unpublished) (holding coworker harassment outside workplace not
actionable under Title VII); Schwapp v. Town of Avon, 118 F.3d 106, 111 (2d Cir. 1997)
(finding that derogatory statements made outside the workplace by plaintiff’s superiors were
relevant to plaintiff’s hostile work environment complaint); Candelore v. Clark County
Sanitation Dist., 975 F.2d 588, 590 (9th Cir. 1992) (excluding evidence of sexual harassment by
coworker that took place outside work hours); Bundy v. Jackson, 641 F.2d 934, 940 n.2 (D.C.
Cir. 1981) (finding that supervisor’s call to plaintiff at her home was considered as contributing
to hostile environment in conjunction with abusive speech at workplace); Blakey v. Continental
Airlines, Inc., 751 A.2d 538, 549 (N.J. 2000) (holding that “harassment by a supervisor that
takes place outside of the workplace can be actionable”) (citing Am. Motorists Ins. Co. v. L-CA Sales Co., 713 A.2d 1007 (N.J. 1998)). But see Darland v. Staffing Res., Inc., 41 F. Supp. 2d
635, 637–39 (N.D. Tex. 1999) (excluding consideration of supervisor’s allegedly harassing
behavior outside the workplace). This consideration by courts of nonworkplace conduct under
406
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
perspective is applied. Much female behavior, and the operation of sexual
selection, is influenced by females’ need or desire for resources.271 Many
male behaviors, as well as physical characteristics, are selected on the basis
of their supposed contribution to the male’s ability to provide resources.272
Indeed, male competition for access to reproductive females and female
need for resources are arguably the two most fundamental assumptions that
drive evolutionary theories of sexual selection.273 These issues likewise are
central to understanding workplace harassment.274
Title VII has been criticized by some scholars as contributing to employer censorship of
employee speech and improper employer control of employees’ private lives. Compare Eugene
Volokh, What Speech Does “Hostile Work Environment” Harassment Law Restrict?, 85 GEO.
L.J. 627, 628–29 n.6 (1997) (noting that the definition of actionable harassment in no way
requires harassing speech to occur within the workplace, and that speech entirely outside the
working environment and working hours can and has been used to support a finding of liability
under Title VII), with J.M. Balkin, Free Speech and Hostile Environments, 99 COLUM. L. REV.
2295, 2312–14 (1999) (explaining that the modern workplace is no longer reducible to a
geographically discrete locale, and that liability for harassing speech may easily result in
censorship well beyond the traditional conception of the “workplace”), and Beverly Earle &
Anita Cava, The Collision of Rights and a Search for Limits: Free Speech in the Academy and
Freedom from Sexual Harassment on Campus, 18 BERKELEY J. EMP. & LAB. L. 282, 299 (1997)
(noting First Amendment implications of campus speech codes proscribing sexual harassment
outside the workplace).
271. E.g., Patricia Adair Gowaty, Sexual Dialectics, Sexual Selection, and Variation in
Reproductive Behavior, in FEMINISM AND EVOLUTIONARY BIOLOGY, supra note 30, at 351–56
(explaining received wisdom of sexual selection theory as including idea that females in most
mammalian and bird species are the limiting resource for males, and that access to biotic and
abiotic resources is a limiting resource for females); Roman Martin Wittig & Christophe
Boesch, “Decision-Making” in Conflicts of Wild Chimpanzees (Pan Troglodytes): An Extension
of the Relational Model, in BEHAVIORAL ECOLOGY AND SOCIOBIOLOGY (2003) (“According to
the socio-ecological model, food is proposed as the most beneficial resource for females
whereas a sexual partner holds the most benefit for a male”); Buss, supra note 138, at 14
(summarizing evolutionary theory that “[f]emales appear to have been limited in reproductive
success by access to resources for self and offspring,” whereas “[m]ales appear to have been
limited by access to fertile females”); Tang-Martinez, supra note 30, at 131 (discussing the
“near dogmatic assumption made by many evolutionary biologists, including most
sociobiologists and many sociobiological feminists . . . that female reproductive success is
limited by a female’s ability to gain access to resources controlled by males, whereas male
reproductive success is limited by access to females.”).
272. This latter generalization only applies to species in which the male contributes more
than just genetic material to the production of offspring. In other species, males provide little or
no resources apart from genetic material. In certain “lekking” species, for example, the male
role in reproduction is solely to contribute sperm. In those species, sexual selection operates to
create males with impressive physical characteristics, but these are understood as a way of
signaling their genetic superiority to females, who look them over in their lek (a large gathering
of males) and then choose one (or more) with which to mate. See RIDLEY, supra note 2, at 140–
42; WILSON, supra note 32, at 331–34.
273. These basic drives similarly are central to an economic explanation of sexuality. See
generally POSNER, supra note 4. Posner posits that the increasing access of females to economic
37:0321]
SEXUAL HARASSMENT LAW
407
The notion that, in humans and other pair-bonding species in which
paternal investment in offspring is significant, a female chooses a male
partly on the basis of his perceived ability to provide resources to her and
her offspring is fundamental to much evolutionary thinking and to much
scholarship on evolutionary sex differences.275 In species, including
humans, in which males often provide some significant parental investment,
“females should seek to mate with males who have the ability and
willingness to provide resources, shelter, territory, and protection.”276
resources independent of their reliance on male support has been the driving force in the social
evolution of sexual relations and family structure. See id.
274. That many victims of workplace sexual harassment are dependent on their jobs for
economic survival has been emphasized in a wide variety of literature on sexual harassment.
See, e.g., MACKINNON, supra note 189, at 41–44 (noting that economic dependence on a
particular job may deter some women from asserting valid sexual harassment claims); Rosa
Ehrenreich, Dignity and Discrimination: Toward a Pluralistic Understanding of Workplace
Harassment, 88 GEO. L.J. 1, 46–47 (1999) (same); Phoebe A. Morgan, Risking Relationships:
Understanding the Litigation Choices of Sexually Harassed Women, 33 LAW & SOC’Y REV. 67,
67–92 (1999) (same). Title VII was in part driven by a recognition that political equality is tied
to economic equality, and that discrimination in employment is thus a serious impediment to
full social and political equality. See H.R. REP. No. 88-914, 88th Congress, at 2402 (Nov. 20,
1963). However, some analyses of sexual harassment emphasize the uniqueness of the
employment context, see, e.g., Cynthia L. Estlund, Freedom of Expression in the Workplace
and the Problem of Discriminatory Harassment, 75 TEX. L. REV. 687, 718–20 (1997)
(proposing the workplace as a unique intermediary between private and public life, and as such
an ideal context for instilling and practicing civic and democratic norms), whereas others
emphasize the more generalized nature of typical harassing behavior in the broader culture, see
MACKINNON, FEMINISM UNMODIFIED, supra note 99, at 103–16. See also Anita Bernstein, An
Old Jurisprudence: Respect in Retrospect, 83 CORNELL L. REV. 1231, 1232–33 & n.14
(engaging Professor Kathryn Abrams on the issue of the centrality of the workplace as a site of
sexual harassment). But see Abrams, supra note 192 (criticizing Bernstein on this count); cf.
Balkin, supra note 270 at 2310 (arguing that “captive audience doctrine” in first amendment
jurisprudence applies to workplace context to permit censorship of employee speech that might
otherwise raise first amendment concerns); Marcy Strauss, Redefining the Captive Audience
Doctrine, 19 HASTINGS CONST. L.Q. 85, 86–87 (1991) (same). An evolutionary perspective
suggests that the workplace context is, indeed, crucial to understanding harassing behavior and
its harms.
275. E.g., Buss, supra note 138, at 1 (1989) (describing a cross-cultural study finding that
women are on average more likely to seek men with higher earning capacity).
276. Id. at 2. There is evidence that calls this evolutionary premise into question. For
example, more than two decades ago evolutionary biologist Patricia Adair Gowaty
demonstrated, in a groundbreaking study, that females of a socially monogamous, pair mating
species of bird fared no differently, reproductively speaking, when they had no access to
resources provided by males than when they did have the help of males. Professor Gowaty
removed the males from her study group and compared their success to that of females in the
control group. She found no statistically significant difference. See Patricia Adair Gowaty, Male
Parental Care and Apparent Monogamy Among Eastern Bluebirds (Sialia sialis), 121 THE AM.
NATURALIST 149, 155–56 (1983). However, these results have varied from species to species
and have been highly controversial. See also ANGIER, supra note 141, at 302–03 (describing
408
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
Several secondary hypotheses follow from this basic premise of sexual
selection theory. First, where females are better able to obtain resources
independent of male investment, they might be expected to make different
reproductive and mate choices than they would if more dependent on males
for resources.277 Second, males who are likely to benefit from female
dependence might be inclined to resist female attempts to become
economically independent. Both of these hypotheses suggest that the
workplace is likely to be a forum of heightened intersexual competition.278
Animal studies of sexually harassing behavior279 demonstrate that the
effects of harassment are more severe when the female is harassed during
resource-gathering activities. Further, these studies suggest that the
behavior is more “successful” (in the sense that the harasser achieves his
supposed objective of sexual access to the female) when the harassment is
connected to a needed resource. In addition, females may “consort” with
dominant males in order to discourage harassment by younger and
subordinate males.280
recent anthropological studies demonstrating that females in hunter-gatherer societies account
for up to 70% of the group’s food).
277. The issue of female choice, and constraints thereon, is a central focus of recent study
and discussion in the evolutionary literature. See infra note 300 and accompanying text.
278. Professor Browne has similarly noted the potential for male-female conflict in the
workplace. He views this conflict as arising not out of female competition with males for
resources which in turn affects male competition for female attention, but instead from the
integration of women into formerly all-male status hierarchies. In his view, a “side-effect of the
breakdown of the sexual division of labor is the expansion of opportunities for sexual conflict to
occur in the workplace.” Browne, Seeking Roots, supra note 4, at 6; see also Browne, Women at
War, supra note 4, at 170–78 (suggesting that the presence of (even individually qualified)
women in combat units will disrupt male bonding and cause sexual conflict). Many feminists,
too, locate the source of female subordination in lack of access to, or control of, resources. E.g.,
Judith Buber Agassi, Theories of Gender Equality: Lessons from the Israeli Kibbutz, in THE
SOCIAL CONSTRUCTION OF GENDER 313, 314–15 (Judith Lorber and Susan A. Farrell eds., 1991)
(summarizing socialist, Marxist, and modern materialist feminist theories). However, these
theories reject an evolutionary or biological explanation for the phenomenon, presumably at
least in part because they equate a biological or genetic explanation with inevitability.
279. Note that ethologists studying animal behavior typically define “sexual harassment” as
unwanted sexual behavior rather than as harassment that occurs “because of” the sex of the
victim. Thus, the scope of the harassing behaviors relevant to these studies is both broader (any
sexual behavior that appears to be unwanted or coercive) and narrower (only sexual behavior)
than that encompassed by “sexual harassment” under Title VII. In practice, most sexual
harassment cases are premised on sexual behavior, see Juliano & Schwab, supra note 6, at 586–
88 (2001), but this may be an effect of the courts’ treatment of these cases and social
understandings of harassment rather than the incidence or harmfulness of sexual as opposed to
nonsexual but sexist harassment. See generally Schultz, supra note 4 (criticizing courts for
failure to consider nonsexual but sexist harassment in hostile work environment cases).
280. See ElizaBeth A. Fox, Female Tactics to Reduce Sexual Harassment in the Sumatran
Orangutan (Pongo pygmaeus abelii), 52 BEHAV. ECOL. SOCIOBIOL. 93, 93–95 (2002); see also
37:0321]
SEXUAL HARASSMENT LAW
409
Thus, a standard premise of evolutionary theory offers a plausible factual
basis for the argument that sexual harassment (and indeed other forms of
treatment that have the intent or effect of driving women from the
workplace or resigning them to low-pay, dead-end jobs) is fundamentally
tied to competition for economic resources. What the evolutionary account
adds to the traditional economic and social accounts is its linkage of
sexually harassing behavior to these economic ends, and its ultimate
grounding of that behavior in evolutionary adaptation. Rather than a private,
noneconomic behavior, sexual harassment in the workplace is revealed,
through an evolutionary lens, to be a type of behavior that may be done by
males for the (evolutionary, ultimate) purpose of restricting females’ access
to resources and thus manipulating females for their own reproductive ends.
Insofar as females, the limiting resource for males as posited by
conventional evolutionary accounts, may be expected to respond to this
“manipulation-control” by males, they will be “selected to remain in control
of their own reproduction.”281 If doing so means striving for economic
independence, then selection pressures should not be expected to operate to
have produced females who forego economically rewarding work by their
own choice, as some have argued.282 Instead, females should be expected to
strive for economic independence against even the most difficult odds. By
many accounts, this is precisely what women have done and continue to
do.283
Mesnick, supra note 253, at 207–208 (advancing the “bodyguard hypothesis” to argue that
females often enter into pair-bonds with males in order to avoid harassment and likely injury by
other males). This finding complicates any claim of “welcomeness” in sexual harassment
doctrine.
281. Gowaty, supra note 65, at 372. For a fascinating account of natural selection in action
in the present day, see generally JONATHAN WEINER, THE BEAK OF THE FINCH (1994)
(describing the work of Peter and Rosemary Grant studying “Darwin’s Finches” in the
Galapagos Islands from the 1970s until the present).
282. Kingsley Browne and Richard Epstein make this argument. See BROWNE, BIOLOGY AT
WORK, supra note 4, at 137 (“If women are less willing than men to sacrifice family for career
and therefore elect not to do so, their decisions might be considered free choices raising no
serious public-policy question.”); Richard A. Epstein, Some Reflections on the Gender Gap in
Employment, 82 GEO. L.J. 75, 80 (1993) (“The key element in all cases is the choice of career
and the willingness to invest in the education and hard work that makes advancement
possible.”); Epstein, Liberty, Patriarchy, supra note 4, at 105–06 (“Knowing what we do about
human nature, do we think that women and men will, in the aggregate, follow similar career
paths in a market system that is not tainted with duress, fraud, or incompetence? I think that the
answer has to be no.”).
283. Evolutionary psychologists often speak of human evolution as if it is a historical
phenomenon rather than an ongoing process. Thus, they focus on the “Environment of
Evolutionary Adaptativeness” as the environment to which the most significant features of
humans should be adapted. While it does make sense to consider human traits in the context of
the environment that is thought to have existed for a large proportion of human evolution, there
410
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
V. DOCTRINAL SPANDRELS IN THE LAW OF SEXUAL HARASSMENT
Title VII prohibits discrimination against an individual on the basis of
that individual’s sex. This much is clear; this much might be viewed in
evolutionary metaphor as the ultimate functional aim of Title VII in the
arena of sex, as locomotion is the ultimate functional aim of human legs.284
There might have been any number of designs that would have
accomplished the same function, some perhaps more efficiently and some
with more elegance. Yet structural and ecological constraints, and perhaps
some degree of chance, have dictated that we humans walk on two legs and
not four, propel ourselves through life on legs and not wheels, move along
the earth and not soar through the air.
Likewise, structural and ecological285 factors have operated to constrain
the development of the law of sexual harassment. As the law prohibiting
sexual harassment in employment has evolved within this legal ecological
niche, several doctrinal spandrels have resulted. These spandrels—these
doctrinal byproducts of other functional aspects of the law—have led to
confusion and needless complexity in the application of sexual harassment
law. In addition, they have at times been mistakenly viewed as central to the
legal design project when in fact they should more appropriately be
understood as secondary. They have been viewed as the arches and not the
spandrels. An understanding of these aspects of doctrine as spandrels and
not arches allows a critical re-evaluation of their usefulness without fear
that the building might come tumbling down.
In this Part, I identify and analyze three “legal spandrels” in sexual
harassment law. The use of the word “sex” in Title VII, combined with the
use of the word “sexual” in harassment law, results in the first spandrel.
Because of the varying linguistic and social meanings that attach to the
words “sex” and “sexual,” the doctrine from the beginning has been rife
is no basis for the conclusion that evolutionary adaptation can no longer occur. If, as current
biological studies suggest, males and females of a species coevolve (and if sexual selection is
driven in large part by immunological factors that are constantly changing), then it also makes
sense to consider current behavior in adaptationist terms.
284. As already noted, see supra note 63, such purposive language is Panglossian in the
context of evolution; it is, however, an oft-used conceptual shortcut.
285. Ecology is “[t]he study of the interrelationships between organisms and their natural
environment, both living and nonliving.” OXFORD DICTIONARY OF BIOLOGY (4th ed. 2000).
When I speak of a legal ecology, I mean to refer to all of the following: the interrelationship of a
particular law or legal doctrine with other laws; the institutional actors involved in the
enactment, enforcement, and interpretation of the law; the language in which the law is
expressed; larger cultural understandings of the law and its policies; and the overall legal and
social discourses within which the law is situated. Others have employed the metaphor of
ecology to law. See, e.g., Peter Manus, One Hundred Years of Green: A Legal Perspective on
Three Twentieth Century Nature Philosophers, 59 U. PITT. L. REV. 557, 569 (1998).
37:0321]
SEXUAL HARASSMENT LAW
411
with ambiguity. I argue that the secondary meanings entailed by these
words have often been mistaken for the central functions of the language
when in fact they should more properly be understood as peripheral.
Second, the birth of sexual harassment law from two very disparate
parents—hostile race environment harassment on the one hand and the
radical feminist critique of sex-as-subordination on the other—has led to the
creation of another legal spandrel. Finally, and most important, the language
of Section 703 of Title VII, which provides that discrimination is actionable
if it is “because of” the victim’s sex, has combined with prevailing
ideologies concerning the contours of intentional discrimination to create a
third legal spandrel. I address each of these areas in turn and argue, based in
part on the evolutionary analysis above, for a refocusing on the primary,
functional aspects of sexual harassment law and away from the distracting,
dysfunctional legal spandrels.
Focusing on the functional, designed elements of sexual harassment law,
I then argue that an evolutionary perspective is wholly consistent with the
position that the three broad patterns of behaviors described and analyzed in
the previous Part lead precisely to the social harms at which Title VII is
aimed. Indeed, once the focus shifts from the sexuality vel non of the
harassing behavior and from the sexual desire or intention of the individual
harasser, the core doctrine that remains is well suited to address the critical
question of whether an individual has been denied an employment
opportunity and has been subjected to discriminatory working conditions
because of her sex.
Furthermore, the evolutionary approach raises the possibility that the
gender blindness of current doctrine may be misplaced.286 Evolutionary
explanations of the patterns of behavior generally at issue in hostile
environment sexual harassment cases reveal that these behaviors are
strongly influenced by the sex of the persons involved. Although this
insight does not mean that behavior toward males should never be
considered sexual harassment sex discrimination, it does suggest that
situations involving male victims might require courts to ask a different set
of questions than those asked when the victim is a woman.
Finally, the analysis reveals that arguments concerning the proper role of
causation in sexual harassment jurisprudence cannot resolve the difficult
cases in which the conscious motivation of the individual harasser is
286. Cf. Kathleen M. Sullivan, Constitutionalizing Women’s Equality, 90 CAL. L. REV. 735,
750 (2002) (discussing the choice that would have to be made by “hypothetical feminist
constitution drafters” between symmetry and asymmetry; that is, between a provision that
would ban discrimination based upon a forbidden classification, or one that would ban
discrimination against a protected class).
412
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
unclear. Rather, the long lens of evolution reveals that the kinds of
behaviors typically involved in harassment cases may almost always be said
to be “based on” the sex of the respective actors; the question of where to
draw the “because of” line is, like the question of proximate cause in tort
law, at bottom, one of social policy and not of factual causation.287 Here,
Supreme Court precedent, the goals of Title VII, and logic all argue for
laying the mantle of intention at the feet of the Title VII defendant—the
employer—and not of the individual harasser.
A. The Terminology Spandrel: What is “Sex?”
The words “sex” and “gender” in discrimination jurisprudence have led
to much confusion.288 First, “sex” may be contrasted with “gender” as a way
of describing certain purported qualities of being male and being female.
Many feminists use the terms to distinguish anatomical maleness and
femaleness (“sex”) from socio-cultural norms regarding the traits generally
associated with masculinity and femininity (“gender”).289 Courts, however,
287. E.g., Leon Green, Proximate Cause in Texas Negligence Law, 28 TEX. L REV. 755,
773–74 (1950). See generally FRANK J. VANDALL ET AL., TORTS: CASES & PROBLEMS (2d ed.
2003).
288. Disagreements over terminology have also led to pitched ideological battles. Some
cultural theorists view language and naming as both reflective and constitutive of social
hierarchies of domination and subordination. See, e.g., JUDITH BUTLER, EXCITABLE SPEECH: A
POLITICS OF THE PERFORMATIVE 27 (1997) (stating that “[t]he utterances of hate speech are part
of the continuous and uninterrupted process to which we are subjected, an on-going
subjection . . . that is the very operation of interpellation, that continually repeated action of
discourse by which subjects are formed in subjugation”). In contrast, Professor Steven Pinker
argues that language is not a “prisonhouse of thought,” but rather is separate from thoughts and
attitudes. PINKER, supra note 31, at 210. He notes that “people invent new words for
emotionally charged referents, but soon the euphemism becomes tainted by association, and a
new word must be found, which soon acquires its own connotations, and so on.” Id. at 212.
Furthermore, purely definitional problems are easily soluble, so that persistent issues of wordmeaning signal deeper substantive conflicts. See Gray, supra note 31, at 387–88. Gray points
out the problem of multiple meanings in the context of biological accounts of the concept of
“innate behavior,” and notes that “[i]f the problem were merely one of sorting out the confusion
caused by using a word with many different meanings, then the issue would have been cleared
up years ago with a few new terms and redefinitions.” Id. at 387.
289. The notion of sex and gender as distinct categories was popularized by “sexologists”
John Money and Anke A. Ehrhardt in their 1972 book MAN & WOMAN, BOY & GIRL: THE
DIFFERENTIATION OF DIMORPHISM OF GENDER IDENTITY FROM CONCEPTION TO MATURITY. They
were concerned with the distinctions between anatomical, chromosomal, and hormonal sex on
the one hand, and psychological gender identity on the other. See ANNE FAUSTO-STERLING,
SEXING THE BODY: GENDER POLITICS AND THE CONSTRUCTION OF SEXUALITY 3–5 (2000)
(discussing the sex/gender dichotomy). Feminists built on this categorical distinction to focus on
the use of social power and dominance to enforce gender identities. See id. at 3 (“[T]he secondwave feminists of the 1970s also argued that sex is distinct from gender—that social
37:0321]
SEXUAL HARASSMENT LAW
413
often use the terms interchangeably, complicating any analysis of what
precisely is prohibited by “sex” discrimination.290 Further muddying the
waters is the dual sense of the word “sex” itself. “Sex” may mean male or
female, but it also refers to the act of engaging in sexual intercourse.291
Thus, sex as a noun must be distinguished from sex as a verb.292 It has been
noted that imprecision in the use of the words “sex” and “gender” has led to
institutions, themselves designed to perpetuate gender inequality, produce most of the
differences between men and women.”)
290. E.g., Harris v. Forklift Sys., Inc., 510 U.S. 17, 21–22 (1993) (“Title VII . . . makes it
‘an unlawful employment practice for an employer . . . to discriminate against any individual
[based on] . . . sex . . . .’”(quoting 42 U.S.C. § 2000e-2(a)(1)). Severe and pervasive conduct
that creates “a work environment abusive to employees [based on] their . . . gender . . . offends
Title VII’s broad rule of workplace equality.”); Hayut v. State Univ. of N.Y., 352 F.3d 733, 748
(2d Cir. 2003) (discussing discrimination based on “sex,” and making multiple references to the
plaintiff being targeted because of her “gender”); McCown v. St. John’s Health Sys., Inc., 349
F.3d 540, 543 (8th Cir. 2003) (referring to “sex” and “gender” interchangeably throughout the
opinion, for example, stating that one evidentiary route of proving same-sex harassment is
through showing motivation of general hostility of people of the same gender in the workplace);
see also Franke, supra note 6, at 712–25. It is apparent, however, that the prohibition against
discrimination “because of . . . sex” in Title VII was meant to refer to sex as a biological
category rather than sexual behavior or sexual orientation. See Simonton v. Runyon, 232 F.3d
33, 36 (2d Cir. 2000) (holding that “the term ‘sex’ in Title VII refers only to membership in a
class delineated by gender, and not to sexual affiliation”); DeCintio v. Westchester County
Med. Ctr., 807 F.2d 304, 306 (2d Cir. 1986) (holding that the word “sex” in Title VII “logically
could only refer to membership in a class delineated by gender, rather than sexual activity
regardless of gender”). In addition, all of the federal courts of appeals that have considered the
question have held that discrimination on the basis of sexual orientation alone does not fall
under Title VII’s proscription against discrimination because of “sex.” See Higgins v. New
Balance Athletic Shoe, Inc., 194 F.3d 252, 259 (1st Cir. 1999); Wrightson v. Pizza Hut of Am.,
Inc., 99 F.3d 138, 143 (4th Cir. 1996); Williamson v. A.G. Edwards & Sons, Inc., 876 F.2d 69,
70 (8th Cir. 1989); DeSantis v. Pac. Tel. & Tel. Co., 608 F.2d 327, 329–32 (9th Cir. 1979).
Some federal appellate courts have been receptive to the argument that discrimination against a
male on the basis of failure to conform to prevailing gender norms can satisfy Title VII’s
requirement of discrimination because of sex. See Simonton, 232 F.3d at 37 (“We find this
argument more substantial than Simonton’s previous two arguments, but not sufficiently pled in
this case.”); Schwenk v. Hartford, 204 F.3d 1187, 1202 (9th Cir. 2000) (stating that “‘sex’ under
Title VII encompasses both sex—that is, the biological differences between men and women—
and gender,” and that harassment based on the victim’s failure to act like a man states a claim
under Title VII); see also Price Waterhouse v. Hopkins, 490 U.S. 228, 258 (1989) (holding that
Title VII was violated where female plaintiff was not promoted in part because she was
insufficiently “macho”); Case, supra note 6, at 36–75; Franke, supra note 6, at 762–71.
291. See MACKINNON, supra note 189, at 182, quoted in Haynie v. Michigan, 664 N.W. 2d
129, 148 n.10 (Mich. 2003) (Cavanagh, J., dissenting) (“It is no accident that the English
language uses the term sex ambiguously to refer both to gender status (as in ‘the female sex’)
and to the activity of intercourse (as in ‘to have sex’).”) (emphasis omitted).
292. Ironically, it was this latter meaning of the word “sex” that might have led to some of
the confusion. Ruth Bader Ginsberg apparently chose to use the term “gender” rather than “sex”
in arguing her groundbreaking discrimination cases because of the unsavory, or titillating,
connotations of the latter word. See Case, supra note 6, at 9–10.
414
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
doctrinal confusion at best and reification of insidious stereotypes at
worst.293 And the use of the word “sex” in harassment jurisprudence has
tended to constrain judicial imagination of the multiple ways that people,
and particularly women, can be subjected to hostile or abusive working
environments.294 On the other hand, opponents of antidiscrimination laws
have objected to the substitution of the word “gender” for “sex” for the
converse reason: the use of the word “gender” has normative implications
about the moral relevance of sex differences that tend to support the
political argument in favor of antidiscrimination laws.295
The introduction of behavioral and evolutionary biology as a
supplemental lens through which to view these issues further complicates
293. See Schultz, supra note 4, at 1704–10; Schwartz, supra note 11; see also Case, supra
note 6, at 2–5 (arguing that much harassment can and should be understood as enforcement of
traditional gender roles); Franke, supra note 6, at 696–97 (arguing that, with the possible
exception of pregnancy rules, it is nonsensical to understand sex discrimination as anything
other than gender—or sex role conformity—discrimination);.
294. The federal courts of appeals have been divided and inconsistent on the issue of
whether nonsexual behavior can constitute harassment on the basis of sex. See Andrews v. City
of Philadelphia, 895 F.2d 1469, 1485 (3d Cir. 1990) (reversing district court “[t]o the extent that
the court ruled that overt sexual harassment is necessary to establish a sexually hostile
environment”); Hall v. Gus Constr. Co., Inc., 842 F.2d 1010, 1014 (8th Cir. 1988) (noting that
“[i]ntimidation and hostility toward women because they are women can obviously result from
conduct other than sexual advances”); Hicks v. Gates Rubber Co., 833 F.2d 1406, 1415 (10th
Cir. 1987) (holding that nonsexual threats of physical violence and racial slurs were properly
considered in support of plaintiff’s sexual harassment claim). But see DeAngelis v. El Paso
Mun. Police Officers Ass’n, 51 F.3d 591, 595–96 (5th Cir. 1995) (finding misogynistic
comments about plaintiff sexist and disparaging, but not overtly sexual); King v. Bd. of Regents
of the Univ. of Wis. Sys., 898 F. 2d 533, 540 (7th Cir. 1990) (separating nonsexual from sexual
elements of plaintiff’s harassment claim, despite her allegation that the combination of such
elements “caused her to be permanently psychologically disabled”); see also Haynie v.
Michigan, 664 N.W. 2d 129, 135–36 (2003) (holding that, under Michigan Civil Rights Act, a
claim of sexual harassment cannot be predicated solely on nonsexual conduct); Juliano &
Stewart, supra note 6, at 555 (finding, based on a comprehensive examination of federal sexual
harassment litigation for the ten-year period following Meritor, that “courts have failed to . . .
acknowledge harassment premised upon nonsexual behavior”). A recent article defines “hostile
environment” harassment as that “consisting of verbal, physical, or environmental behavior that
is sexual in nature and has the effect of creating a hostile, offensive, or abusive working
environment.” Joanna L. Grossman, The Culture of Compliance: The Final Triumph of Form
over Substance in Sexual Harassment Law, 26 HARV. WOMEN’S L.J. 3, 10 n.29 (2003)
(emphasis added).
295. See Epstein, Nouns, supra note 145; cf. J.E.B. v. Alabama ex rel. T.B., 511 U.S. 127,
157 n.1 (1994) (Scalia, J., dissenting). For feminist arguments that courts should be careful of
using the word “sex” when they mean “gender,” and vice versa, see Case, supra note 6, at 12
(arguing that “gender” is more appropriately used as an adjective, describing “masculine” or
“feminine”); Franke, Disaggregation, supra note 192, at 1–9. But see Franke, supra note 6, at
696 (arguing that the distinction is faulty and insidious to the cause of equality—basically
arguing that there is no such thing as “sex”; everything is “gender”).
37:0321]
SEXUAL HARASSMENT LAW
415
the definitional problems. Though some feminists use the term “sex” to
refer to physical and immutable differences between males and females
(such as the biological fact that only females can become pregnant, give
birth, and lactate), in biology the line between genes and environmental
influences is difficult or impossible to draw.296 In particular, scientists who
study the biological influences on human and nonhuman animal behavior
would consider many of the so-called “gendered” behavioral patterns as at
least partly biological or genetic in origin, and vice versa.297 Oxford
biologist Richard Dawkins refers to these kinds of behaviors as “the
extended phenotype,”298 in order to emphasize that the influences of the
genes can reach beyond the physical appearance of the individual organism,
and even beyond the behavior of the individual organism, to affect entire
ecosystems.299 Indeed, selection works directly upon phenotype, not
genotype; phenotype plasticity is of major practical and theoretical
importance in the current study of evolutionary biology.300
The behavioral biological perspective, therefore, makes the distinction
between sex-as-physiology and gender-as-cultural-norm virtually
incoherent. Discrimination is necessarily relational. When a man harasses a
woman “because” she is a woman, he does so based upon some interaction
296. In fact, most scientists believe that there is no line, but rather a sort of permeable
membrane through which genetic and environmental influences pass back and forth and act
upon one another in uncounted ways. See PINKER, supra note 31, at 60; RIDLEY, supra note 31,
at 278 (“Each person is molded by an interaction of his environment, especially his cultural
environment, with the genes that affect social behavior.”) (quoting E. O. Wilson); WESTEBERHARD, supra note 31, at 33 (“To solve the nature-nurture problem, one has to acknowledge
the deterministic role of the environment, alongside the genes, in development and to consider
the development of the entire phenotype.”).
297. “Separating sex from gender does not allow for the ways in which experiential inputs
shape physiology and anatomy, and the ways in which physiological and anatomical differences
shape behavior and experience.” Gray, supra note 31, at 406 (citing R. HUBBARD, THE POLITICS
OF WOMEN’S BIOLOGY (1990)).
298. Phenotype is defined as “[t]he observable characteristics of an organism. These are
determined by its genes, the dominance relationships between the alleles, and by the interaction
of the genes with the environment.” OXFORD DICTIONARY OF BIOLOGY 453 (4th ed. 2000)
(internal cross-references omitted).
299. See generally RICHARD DAWKINS, THE EXTENDED PHENOTYPE: THE GENE AS THE UNIT
OF SELECTION (1982). For a briefer summary of this idea, see DAWKINS, supra note 37, at 134–
37. Though he does not address the issue of cultural gender roles and behavior, the general
framework of his theory of extended phenotypes would apply to such roles and behavior if they
have a genetic component. Dawkins offers the example of a beaver’s dam to illustrate the ways
in which a genetically-driven behavior can have widespread effects. He views the pond created
by the beaver’s dam, and the consequent interactions of other species in the pond, as the
extended phenotype of the beaver’s gene for dam-building. Id. at 135–36.
300. See WEST-EBERHARD, supra note 31, at vii (offering a “synthesis of development and
evolution” primarily for a scientific audience).
416
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
between her phenotype (the way her femaleness is outwardly expressed
both physically and behaviorally) and his ideas and beliefs about that
“Woman” phenotype.301 He cannot discriminate based directly upon her
genotype, or upon the presence in her cells of an X chromosome, because
those are invisible to him. Sex discrimination, then, is phenotype
discrimination; phenotype discrimination is “gender” discrimination insofar
as the concept of gender encompasses the outward expression of “sex.”
Because discrimination consists of harmful treatment that stems from
one person’s ideas and beliefs about another person’s extended
phenotype,302 it is appropriate to describe and analyze sex discrimination as
gender discrimination. In this sense, a focus on sex-as-biology or on sex-assexual-behavior is a misleading and dysfunctional spandrel in the law of
sexual harassment.
B. The “Mother’s Eyes, Father’s Nose” Spandrel: Who Are the Parents of
Hostile Environment Sexual Harassment Law?
Hostile work environment sexual harassment doctrine was born of two
very different jurisprudential parents: an analogy to hostile racial
environment harassment on the one hand303 and as a sort of adjunct to the
301. Cf. FINEMAN, supra note 15, at 51 (arguing that, notwithstanding debates about
biology, culture, and essentialism, “women will be treated as mothers (or potential mothers)
because ‘Woman’ as a cultural and legal category inevitably encompasses and incorporates
socially constructed notions of motherhood in its definition”). When this insight is applied to
treatment of women on an individual level, it suggests that particular actors will treat particular
women based in part upon their ideas of “Woman,” however acquired.
302. There is abundant evidence that animals make relational choices based upon
phenotype. For example, in a series of studies biologists found “that male swallows with
artificially lenghtened [sic] tails acquired mates more quickly, reared more young, and had more
adulterous affairs than males of normal length.” RIDLEY, supra note 3, at 137 (citing A.P.
Møller, Female Choice Selects for Male Sexual Tail Ornaments in the Monogamous Swallow,
332 NATURE 640, 640–42 (1988)). Interestingly, scientists have also found that the more
“attractive” males are significantly less helpful around the nest. See id. at 224.
303. See Henson v. City of Dundee, 682 F.2d 897, 902 (11th Cir. 1982) (noting that
“[s]exual harassment which creates a hostile or offensive environment for members of one sex
is every bit the arbitrary barrier to sexual equality at the workplace that racial harassment is to
racial equality”); Walker v. Ford Motor Co., 684 F.2d 1355, 1358 (11th Cir. 1982) (applying the
legal standard enunciated in Henson to claims for hostile work environment based on either race
or sex); Bundy v. Jackson, 641 F.2d 934, 945 (D.C. Cir. 1981) (explaining the relevance of
“discriminatory environment” cases to sexual harassment). Of course, both race and sex
discrimination may be present in the same case. The particular problems faced by victims at the
crossroads of multiple subordinating identities have recently been the focus of several scholarly
works. See generally Crenshaw, supra note 94, at 150–67; Harris, supra note 94, at 590–605.
While I recognize that these negative synergies of discrimination complicate any analysis of
sexual harassment law, here I focus only on sex, and not race, discrimination.
37:0321]
SEXUAL HARASSMENT LAW
417
paradigm quid pro quo harassment case, embodying the feminist conception
of male use of sex and sexuality as a method of subordinating women, on
the other.304 When this offspring was then raised in an environment in which
the word “sex” is susceptible of several meanings, doctrinal confusion and
legal spandrels were bound to result.
1.
The Analogy to Race
The first federal case to recognize a claim of hostile work environment
harassment arose in the context of a Title VII claim of racial discrimination.
In Rogers v. EEOC,305 the plaintiff claimed that her employer, an optometry
practice, engaged in “patient segregation” whereby it treated or classified
patients depending upon their race. The court interpreted this part of the
plaintiff’s claim to allege that the practice was discriminatory towards her,
and held it sufficient to allow the claim to proceed.306 In a now well-known
passage, the court stated that “the phrase ‘terms, conditions, or privileges of
employment’ in Section 703 is an expansive concept which sweeps within
its protective ambit the practice of creating a working environment heavily
charged with ethnic or racial discrimination.”307 The court thus recognized
that imposing upon a person the requirement that she work under conditions
of extreme racial animosity, bigotry, or ridicule implicated a “term or
condition” of employment sufficient to trigger Title VII protection
regardless of whether a tangible employment action was taken.
Following Rogers, several federal courts of appeals recognized that the
existence of a racially hostile environment, even absent a tangible job
action,308 could give rise to a valid claim of employment discrimination.309
304. See Williams v. Bell, 587 F.2d 1240, 1245–46 (D.C. Cir. 1978) (finding that
retaliatory dismissal for refusal to accede to sexual advances violated Title VII); Tomkins v.
Pub. Serv. Elec. & Gas Co., 568 F.2d 1044, 1045 (3d Cir. 1977) (same); Barnes v. Costle, 561
F.2d 983, 994–95 (D.C. Cir. 1977) (reversing district court’s grant of summary judgment to
employer that conditioned plaintiff’s advancement on acquiescence to supervisor’s sexual
demands); MACKINNON, supra note 189, at 174–92.
305. 454 F.2d 234 (5th Cir. 1971), cert. denied, 406 U.S. 957 (1972).
306. Id. at 240–41.
307. Id. at 238.
308. Because of the potential expansiveness of this premise, courts began right away to
impose various limitations upon the hostile environment claim. One possible limitation,
eventually rejected by the Supreme Court in a nonrace hostile environment case, centered on
requiring that the plaintiff demonstrate some measurable harm. Because by definition the hostile
environment claim did not result in a tangible employment action, by necessity the required
harm was psychological or emotional. See, e.g., Harris v. Forklift Sys., Inc., 976 F.2d 733 (6th
Cir. 1992) (affirming judgment for employer where harassing conduct did not affect the
plaintiff’s psychological well-being), rev’d, 510 U.S. 17, 22 (1993) (finding that the district and
appellate courts erroneously relied on the element of psychological harm, and noting that “Title
418
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
Whereas in Rogers the abusiveness of the environment toward the plaintiffemployee stemmed from the employer’s discriminatory treatment and
classification of the clients of the business, in subsequent cases the
abusiveness of the environment often stemmed from pervasive racist
expression or conduct in the workplace.310 Once courts had held that
offensive racial expression and behavior in the workplace could, without
more,311 state a claim under Title VII, it “follow[ed] ineluctably” that a
hostile work environment caused by sexual harassment could likewise state
a Title VII claim.312
Once the requirement of a tangible harm was discarded, it became
necessary to limit in some alternative way the perceived vast sweep of the
hostile environment cause of action.313 In the hostile environment situation,
VII bars conduct that would seriously affect a reasonable person’s psychological well-being, but
the statute is not limited to such conduct”); Sparks v. Pilot Freight Carriers, Inc., 830 F.2d 1554,
1561 (11th Cir. 1987) (same); Rabidue v. Osceola Ref. Co., 805 F.2d 611, 620 (6th Cir. 1986),
cert. denied, 481 U.S. 1041 (1987) (requiring plaintiff to demonstrate that “sexual harassment . .
. seriously [affected her] psychological well-being”); Downes v. F.A.A., 775 F.2d 288, 292–93
(Fed. Cir. 1985) (same). A second limitation, still extant, requires that the harassment have been
“unwelcome.” E.g., Meritor Sav. Bank v. Vinson, 477 U.S. 57, 68 (1986) (stating that “the
gravamen of any sexual harassment claim is that the alleged sexual advances were
‘unwelcome’”).
309. See Snell v. Suffolk County, 782 F.2d 1094 (2d Cir. 1986); Gilbert v. City of Little
Rock, 722 F.2d 1390 (8th Cir. 1983); Vaughn v. Pool Offshore Co., 683 F.2d 922 (5th Cir.
1982); Johnson v. Bunny Bread Co., 646 F.2d 1250 (8th Cir. 1981); DeGrace v. Rumsfeld, 614
F.2d 796 (1st Cir. 1980); Firefighters Inst. for Racial Equal. v. St. Louis, 549 F.2d 506 (8th Cir.
1977); Gray v. Greyhound Lines, East, 545 F.2d 169 (D.C. Cir. 1976); see also Cariddi v. Kan.
City Chiefs Football Club Inc., 568 F.2d 87 (8th Cir. 1977) (national origin).
310. E.g., Snell, 782 F.2d at 1103 (affirming judgment for plaintiffs where workplace
evidenced proliferation of demeaning literature and epithets and repeated ethnic slurs, in
addition to incident in which Hispanic inmate was humiliated in the presence of Hispanic
officer); Gilbert, 722 F.2d at 1394 (implicitly accepting that widespread use and official
condoning of racial remarks, derogatory epithets, racially-oriented graffiti, and the like could
state a claim of hostile environment discrimination, but upholding as not clearly erroneous
district court finding that the evidence was insufficient to establish a pattern of harassment in
that case); Johnson, 646 F.2d at 1257 (holding that “[u]nquestionably, a working environment
dominated by racial slurs constitutes a violation of Title VII,” but finding in that case that the
racial comments were insufficiently frequent, steady, and directed to violate Title VII).
311. This is not to suggest that there are no further requirements for proving a hostile
environment claim, but rather that in theory such a claim can stem solely from abusive,
insulting, or humiliating words and/or conduct.
312. Bundy v. Jackson, 641 F.2d 934, 943 (D.C. Cir. 1981); see also Henson v. City of
Dundee, 682 F.2d 897, 902 (11th Cir. 1982).
313. In one of the earliest cases to consider a claim of sexual harassment, the district court
(in)famously expressed the sentiment:
If the plaintiff’s view were to prevail, no superior could, prudently,
attempt to open a social dialogue with any subordinate of either sex. An
invitation to dinner could become an invitation to a federal lawsuit if a once
37:0321]
SEXUAL HARASSMENT LAW
419
then, both the race cases and the sex cases have tended to focus on the
offensiveness of words or conduct that so permeate the environment as to
adversely affect the terms and conditions of the plaintiff’s employment.314
In the racial context, offensiveness is relatively unambiguous, and courts
seem content to perform their “gatekeeper” function with reference to the
requirement that the harassment, in order to violate Title VII, be “severe or
pervasive.”315 In the context of harassment based on sex, in contrast, courts
have sometimes viewed the question of offensiveness with greater
ambivalence.
This ambivalence is potentially two-fold. First, there is the perception
that the discriminatory effect of sexual advances, language, and behavior in
the workplace depends both on the subjectivity of the victim316 and on the
harmonious relationship turned sour at some later time. And if an inebriated
approach by a supervisor to a subordinate at the office Christmas party could
form the basis of a federal lawsuit for sex discrimination if a promotion or a
raise is later denied to the subordinate, we would need 4,000 federal trial
judges instead of some 400.
Tomkins v. Pub. Serv. Elec. & Gas Co., 422 F. Supp. 553, 557 (D.N.J. 1976), rev’d, 568 F.2d
1044 (3d Cir. 1977). See also Oncale v. Sundowner Offshore Servs., Inc., 523 U.S. 75, 80
(1998) (explaining that Title VII is not to be used as a “general civility code for the American
workplace,” and that the Court has “never held that workplace harassment . . . is automatically
discrimination . . . merely because the words used have sexual content or connotations”);
Faragher v. City of Boca Raton, 524 U.S. 775, 788 (1998) (reiterating that hostile work
environment standards must be “sufficiently demanding to ensure that Title VII does not
become a ‘general civility code’”).
314. See, e.g., Oncale, 523 U.S. at 81 (noting that Title VII “forbids only behavior so
objectively offensive as to alter the ‘conditions’ of the victim’s employment”); Black v. Zaring
Homes, Inc., 104 F.3d 822, 827 (6th Cir. 1997) (explaining that “there may be times when
offensive comments have an impact in the workplace, indeed constitute ‘harassment,’ but do not
create an objectively hostile work environment”).
315. E.g., Briggs v. Anderson, 796 F.2d 1009 (8th Cir. 1986) (a few instances of racial slurs
being uttered not severe or pervasive); Fortenberry v. United Airlines, 28 F. Supp. 2d 492 (N.D.
Ill. 1998) (insulting racial terms did not rise to the level of a hostile environment claim); Harris
v. SmithKline Beecham, 27 F. Supp. 2d 569 (E.D. Pa. 1998) (racial comment made to AfricanAmerican plaintiff by her supervisor was not severe).
316. To prevail on a hostile environment claim, the plaintiff must prove that the
complained-of conduct was both objectively and subjectively hostile or abusive. Harris v.
Forklift Sys., Inc., 510 U.S. 17, 21–22 (1993). The subjective component involves an inquiry
into whether the particular plaintiff in fact “subjectively perceive[d] the environment to be
abusive,” because if she did not, then “the conduct has not actually altered the conditions of the
victim’s employment, and there is no Title VII violation.” Id. at 22–23. The objective element,
however, has not been uniformly applied by the federal circuit and district courts. Some courts
focus on whether alleged harassment would be viewed as abusive by a “reasonable person.” See,
e.g., Richardson v. N.Y. State Dep’t of Corr. Serv., 180 F.3d 426, 436 (2d Cir. 1999) (applying
a “reasonable person” standard); DeAngelis v. El Paso Mun. Police Officers Ass’n, 51 F.3d 591,
594 (5th Cir. 1995) (same); Hirschfeld v. N.M. Corr. Dep’t, 916 F.2d 572, 580 (10th Cir. 1990)
(same); Steele v. Offshore Shipbuilding, Inc., 867 F.2d 1311, 1317 (11th Cir. 1989) (same);
420
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
manner in which she has projected her subjective reaction.317 Identical
sexualized conduct might constitute harassment in one context and might be
perfectly appropriate in another, according to this reasoning.318 Second,
even assuming that the plaintiff is subjectively offended and that the
conduct is “unwelcome,” it is possible that social norms that disapprove of
racial bigotry operate more strongly than parallel norms against offensive
sexual language and conduct.319
Paroline v. Unisys Corp., 879 F.2d 100, 105 (4th Cir. 1989) (same); Brooms v. Regal Tube Co.,
881 F.2d 412, 414 (7th Cir. 1989) (same). Other courts have tailored the inquiry further still,
employing a “reasonable woman” or “reasonable minority” standard. See, e.g., Burns v.
McGregor Elec. Indus., Inc., 989 F.2d 959, 964 (8th Cir. 1993) (applying a “reasonable woman”
standard); Ellison v. Brady, 924 F.2d 872, 879 (9th Cir. 1991) (same); Andrews v. City of
Philadelphia, 895 F.2d 1469, 1482, 1486 (3d Cir. 1990) (same); Yates v. Avco Corp., 819 F.2d
630, 637 (6th Cir. 1987) (same); Stingley v. Stare, 796 F. Supp. 424, 428 (D. Ariz. 1992)
(applying a “‘reasonable person of the same gender and race or color’ standard”); Harris v. Int’l
Paper Co., 765 F. Supp. 1509, 1516 (D. Me. 1991), vacated in part by 765 F. Supp. 1529 (D.
Me. 1991) (applying a “reasonable black person” standard). There has also been much debate
among academics concerning the appropriate reasonableness benchmark. See, e.g., Nancy S.
Ehrenreich, Pluralist Myths and Powerless Men: The Ideology of Reasonableness in Sexual
Harassment Law, 99 YALE L. J. 1177, 1214–1233 (1990); Saba Ashraf, The Reasonableness of
the “Reasonable Woman” Standard: An Evaluation of its Use in Hostile Environment Sexual
Harassment Claims Under Title VII of the Civil Rights Act, 21 HOFSTRA L. REV. 483 (1992);
Leslie M. Kerns, A Feminist Perspective: Why Feminists Should Give the Reasonable Woman
Standard Another Chance, 10 COLUM. J. GENDER & L. 195 (2001).
317. I refer here to the requirement that the plaintiff demonstrate that the sexual conduct
was “unwelcome.” See Meritor Sav. Bank v. Vinson, 477 U.S. 57, 68 (1986) (stating that “[t]he
gravamen of any sexual harassment claim is that the alleged sexual advances were
‘unwelcome’”)(citing 29 C.F.R. § 1604.11(a) (1985). This requirement has been criticized as
importing into sex discrimination law the same myths and sexual stereotypes that have long
plagued the law of rape. See Estrich, supra note 267, at 826–34.
318. Indeed, this is the central tension in the law of rape. Cf. Estrich, supra note 315
(describing the myriad of ways in which the sexual harassment doctrine replicates the malebiased assumptions in the law of rape). In noting the disanalogy to race, Professor Hébert has
pointed out that it “is difficult to imagine the Supreme Court” imposing an unwelcomeness
requirement on racially harassing conduct or suggesting that a plaintiff’s demeanor might have
invited the harassment. See L. Camille Hébert, Analogizing Race and Sex in Workplace
Harassment Claims, 58 OHIO ST. L.J. 819, 824 (1997). But see Powell v. Misourri State Hwy.
and Transp. Dep’t, 822 F.2d 798, 801 (8th Cir. 1987) (finding that racially charged language did
not rise to the level of actionable harassment where plaintiff participated in banter by calling
white coworkers “honky” and “whitey”); Jones v. First Federal Sav. and Loan Ass’n of
Winston-Salem, 546 F. Supp. 762, 779 (M.D.N.C. 1982) (summary judgement granted in favor
of employer where both blacks and whites engaged in sporadic comments and joking, with
black employees referring to their white coworkers “soda crackers” and “honkies”); Vaughn v.
Pool Offshore Co., 683 F.2d 922, 924 (5th Cir. 1982) (seemingly racially offensive terms “were
bandied back and forth without apparent hostility or racial animus”).
319. See Hébert, supra note 318, at 833 n.54 (“Arguably, the difference is that sexist
behavior is perceived as relatively more acceptable, or as relatively less offensive, than racist
behavior . . .”) (quoting Joshua F. Thorpe, Note, Gender-Based Harassment and the Hostile
Work Environment, 1990 DUKE L.J. 1361, 1396).
37:0321]
SEXUAL HARASSMENT LAW
421
Moreover, there is no clear racial analog to the quid pro quo sexual
harassment case or to the hostile environment case that is predicated upon
unwelcome sexual advances aimed, however inappropriately, at establishing
some sort of sexual relationship with the plaintiff.320 Though the analogies
between race (and color, religion, or national origin) and sex are often apt in
the hostile environment context, they are nonexistent in the quid pro quo
context. There, the sexual nature of the conduct distinguishes it from
virtually all other kinds of actionable discrimination.321 In this respect,
judicial acceptance of the discriminatory nature of sexually coercive
behavior in the workplace emerges as a genuinely radical development.
2.
The Feminist Anti-Subordination Argument
In addition to the analogy to hostile racial environment cases, the judicial
acceptance of sexual harassment as a violation of Title VII was informed by
the radical feminist description of sexuality and power. One of the primary
architects of the doctrine, Professor Catharine MacKinnon, has described
human sexuality largely as both an expression of, and a mechanism for
perpetuating, male domination over women.322 Though this position was
certainly not embraced wholeheartedly by the Supreme Court in Meritor, to
a large extent its assumptions drove the development of the law of sexual
harassment. The paradigmatic sexual harassment case was, and remains,
320. This seems to be one of Justice Thomas’s concerns in his dissent in Ellerth. Burlington
Indus., Inc., v. Ellerth, 524 U.S. 742, 766–67 (1998) (Thomas, J., dissenting). He argues that the
Court’s “rule that employers are vicariously liable if supervisors create a sexually hostile work
environment, subject to an affirmative defense that the Court barely attempts to define” has the
result that “employer liability under Title VII is judged by different standards depending upon
whether a sexually or racially hostile work environment is alleged.” Id. Though the dissent does
not elaborate, it appears to assume that employer liability for hostile environment claims
predicated upon race will continue to be judged under a negligence standard where a tangible
employment action has not been shown.
321. Some scholars have criticized Title VII doctrine and theory for taking insufficient
account of the ways in which race and sex combine and interrelate in many cases of
discrimination. See Abrams, supra note 94, at 2493–2517; Harris, supra note 94, at 585–90. The
argument that racial harassment and sexual harassment are not analogous across the category of
cases in which the harasser is motivated by sexual desire does not imply that race is necessarily
(or even potentially) irrelevant in those cases. It implies only that there is no similar “desire”
category of cases in a pure racial harassment context.
322. In Sexual Harassment of Working Women, Professor MacKinnon defines “sexual
harassment” as “the unwanted imposition of sexual requirements in the context of a relationship
of unequal power. Central to the concept is the use of power derived from one social sphere to
lever benefits or impose deprivations in another.” She goes on to state that sexual harassment is
“one dynamic which reinforces and expresses women’s traditional and inferior role in the labor
force.” MACKINNON, supra note 189, at 1, 4.
422
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
that of a male supervisor abusing his power over subordinate female
employees to extort, or attempt to extort, sexual favors.
In Sexual Harassment of Working Women, first published in 1979,
Professor MacKinnon outlined two broad categories of sexual harassment
discrimination. The first, which she labeled “quid pro quo” harassment, was
“defined by the more or less explicit exchange: the woman must comply
sexually or forfeit an employment benefit.”323 However, MacKinnon
understood the category broadly to include situations in which the woman
complies with the threat.324 She argued that the injury consisted in “being
placed in the position of having to choose between unwanted sex and
employment benefits or favorable conditions.”325
The injury at issue in the quid pro quo case is of a different nature than
the harm that results when a person is subject to a working environment
suffused with severe or pervasive “intimidation, ridicule, and insult.”326
According to the Supreme Court, a hostile environment “can and often will
detract from employees’ job performance, discourage employees from
remaining on the job, or keep them from advancing in their careers.”327 In
addition, the Court has hinted at a kind of dignitary injury328 separate from
these more tangible harms. Holding that a plaintiff could survive summary
judgment even absent a showing of “tangible psychological injury,” the
Court stated that “the very fact that the discriminatory conduct was so
severe or pervasive that it created a work environment abusive to employees
because of their race, gender, religion, or national origin offends Title VII’s
broad rule of workplace equality.”329 Putting aside the hopelessly circular
nature of this reasoning, the impression remains that the injuries that result
from being forced to work in a discriminatorily abusive environment are
distinct from those that result from being pressured by someone in a relative
position of power to engage in a sexual relationship.
323. Id. at 32.
324. See id. at 32–33.
325. Id. at 37.
326. The Court in Meritor characterized hostile environment discrimination in these terms.
See Meritor Sav. Bank v. Vinson, 477 U.S. 57, 65 (1986).
327. Harris v. Forklift Sys., Inc., 510 U.S. 17, 22 (1993).
328. For an argument that sexual harassment liability should be evaluated under a
“respectful person” standard, see generally Anita Bernstein, Treating Sexual Harassment with
Respect, 111 HARV. L. REV. 445 (1997–98). Though her article focuses on the standard to be
applied to determine whether particular behavior satisfies the objective prong of the “hostile or
abusive” element, Professor Bernstein’s analysis treats sexual harassment as a dignitary harm.
Id. at 509 (“That hostile environment sexual harassment is fundamentally an injury to dignity
escapes few who have experienced and studied the phenomenon.”).
329. Harris, 510 U.S. at 22.
37:0321]
SEXUAL HARASSMENT LAW
423
The EEOC and the Supreme Court have blurred the lines between these
conceptually distinct forms of discrimination by referring to both under the
rubric “sexual harassment.” In addition, the Supreme Court has narrowed
the usefulness of the quid pro quo harassment category by defining it
narrowly only to include “threats to retaliate against [a plaintiff] if she
denied some sexual liberties” where those “threats are carried out.”330 In
other cases involving attempts to leverage power for sex, the Court applies
the label hostile work environment sexual harassment. Thus the Court has
coopted these definitional categories to correspond to its formulation of
when an employer is entitled to assert an affirmative defense to vicarious
liability.331
Quid pro quo harassment as originally defined by MacKinnon is
encompassed by the first pattern of behavior analyzed in Part III, in which a
harasser attempts to get sex from the victim. Conduct of coworkers that is
also aimed at coercing sex is also included in the first category. These are
the behaviors with no clear analogue in the race, religion, or national origin
contexts. These are the behaviors that are or should be understood as sexual
as opposed to simply based on sex.
Cabining the term “sexual harassment” to quid pro quo situations,
broadly understood, would liberate hostile environment harassment from
the confusion and ambiguity engendered by the use of the word “sexual” in
connection with the pure hostile environment cases. Once the “sexual” in
hostile environment harassment is replaced by “sex” or “sexist,” or perhaps
“gender,” the misapprehension that sexual behavior is required to create a
hostile environment will more easily be dispelled, and the analogy to the
hostile racial environment cases will be clear.
A redefinition (or, more accurately, relabeling) of the two broad
categories of “harassment on the basis of sex”332 in this way, roughly tracks
330. Burlington Indus., Inc. v. Ellerth, 524 U.S. 742, 751 (1998). The Court’s use of the
term corresponds only to the first of several types of quid pro quo harassment as originally
defined by MacKinnon. See MACKINNON, supra note 189, at 32–40.
331. See infra text accompanying notes 344–351.
332. There are two potential problems with insisting that “sexual harassment” be referred to
as “harassment on the basis of sex.” First, the phrase “harassment on the basis of sex” is rather
awkward; exhorting judges to use a phrase that is both unwieldy and not in common usage
seems bound to fail. Yet the virtues of the phrase in the legal context are substantial: it
encompasses the causation requirement of Title VII; it tracks more closely the terms used in
other categories of discrimination; and it is much less subject to misinterpretation and
misapplication. Second, the phrase “sexual harassment” is widely used and understood in the
popular culture, and restricting that use is neither desirable nor possible. However any problem
posed by this inconsistency is more apparent than real. Even as presently understood, the term
“sexual harassment” has different meanings in the popular and legal contexts. See, e.g., Merit
Systems Protection Board Report on Sexual Harassment (stating that the definition in the survey
424
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
the first two broad classes of cases outlined in Part III. Quid pro quo
harassment is behavior designed to “get sex” from the victim; hostile
environment harassment on the basis of sex is an aggregation of words and
conduct, sexual or not, designed to exclude or control the victim or victims,
often through humiliation or fear. The important distinction between the
two is not whether the harassment includes sexually-explicit or suggestive
words or conduct. Rather, the more useful line is between behavior
designed to solicit or coerce sex and that designed to humiliate, demean, or
expel a person because of that person’s sex or sex role signaling.333 That
sexually explicit language or conduct might be used in the latter context is
relevant only because it is a common strategy of harassment. And while
commentators frequently wonder why women are on average more
offended than men by sexually explicit words and conduct,334 it is
of “sexual harassment” is different from the legal definition); see also sources cited supra note
194.
333. Cf. Fremling & Posner, supra note 4, at 1080–1101 (explaining the harm of sexual
harassment as primarily a status offense). These authors focus on the economic and
evolutionary explanations for “status signaling” behavior in males and females, and argue that
much sexual solicitation is offensive to females because it suggests that they are of lower status
than they actually are or believe themselves to be. At times the authors sound like they inhabit
an alternate universe. “[E]ven if the solicitation is private and others are not aware, the woman’s
self-esteem will still be damaged because she will realize that at least one man doubts whether
she is really a high-status woman.” Id. at 1082. And “in deciding whether to permit nude pinups
in the workplace, the employer would trade off the costs to its female employees (who would
insist on being compensated through higher wages) against the benefits to its male employees,
who would accept lower wages.” Id. at 1088. Nonetheless, their focus on signaling as an
evolutionary strategy is instructive. Humans signal not only status, but also gender; statusenforcing norms operate alongside gender-enforcing norms.
334. Studies have shown that women are more offended by sexually explicit words and by
“curse” words than are men; there is also one intriguing study in which, when people were told
to role-play opposite gender roles, women imagining they were men rated the words as less
offensive than women-as-women. See JAY, supra note 184, at 186–88. Some feminists have
argued that the protectionist tendency of Title VII case law that assumes that sexually-explicit
language or a sexually-charged atmosphere is more offensive to women than to men is itself
sexist and reinforces harmful “repronormal” social stereotypes of women. See Katherine M.
Franke, Theorizing Yes: An Essay on Feminism, Law and Desire, 101 COLUM. L. REV. 181, 201
(2001)
I wonder if an intergenerational moment might have arrived when we would
want to de-sacrilize the sex-danger alchemy within feminist legal theory—
not to ignore the significance of sexual violence for women, but instead to
de-essentialize sex’s a priori status as a site of danger for women and one
best cleansed of such danger.
Id.; cf. NADINE STROSSEN, DEFENDING PORNOGRAPHY: FREE SPEECH, SEX, AND THE FIGHT FOR
WOMEN’S RIGHTS 247–64 (1995) (arguing that censorship of pornography on the ground that it
harms women actually works to undermine women’s equality and autonomy).
37:0321]
SEXUAL HARASSMENT LAW
425
nonetheless not uncommon for men to be harassed by this method as
well.335
C. The Causation Spandrel: Whose Intention Counts?
The “because of” language in the text of Section 702 of Title VII,
combined with an ideological schema that tends generally to locate
discrimination in notions of individual fault and wrongful motives,336 has
resulted in an unfortunate focus in hostile environment sex discrimination
cases upon the intentions and motivations of individual harassers. This
preoccupation with the fault of the individual harasser is a third legal
spandrel in sexual harassment law. In fact, individual fault is at best a
neutral byproduct of the doctrine; insofar as it is relevant to liability, it
should be presumed wherever the harassment fits one of the three broad
patterns described in Part IV, and knowledge that such behaviors occur
“because of . . . sex” should be imputed to the employer—the relevant
causal actor under Title VII.
There are several reasons that a focus on intent is especially problematic
in hostile environment cases. First, though it is the employer who is liable to
335. E.g., Oncale v. Sundowner Offshore Servs., Inc., 523 U.S. 75, 77 (1998) (deciding a
same-sex harassment suit claiming former supervisors and coworkers called plaintiff names
with homosexual overtones); King v. Super Serv., Inc., 68 Fed. Appx. 659, 660 (6th Cir. 2003)
(unpublished opinion) (entertaining but rejecting same-sex harassment claim); Davis v. Coastal
Int’l Sec., Inc., 275 F.3d 1119, 1121 (D.C. Cir. 2002) (deciding a same-sex harassment suit
alleging that coworkers grabbed plaintiff’s crotch, made kissing gestures, and requested oral
sex); Bibby v. Phila. Coca Cola Bottling Co., 260 F.3d 257, 262–63 (3d Cir. 2000) (extending
protection in same-sex cases to pure gender stereotyping by holding that “a plaintiff may be able
to prove that same-sex harassment was discrimination because of sex by presenting evidence
that the harasser’s conduct was motivated by a belief that the victim did not conform to the
stereotypes of his or her gender”); Quick v. Donaldson Co., Inc., 90 F.3d 1372, 1378 (8th Cir.
1996) (acknowledging same-sex cause of action); Hampel v. Food Ingredients Specialties, Inc.,
729 N.E.2d 726, 729 (Ohio 2000) (considering action of plaintiff’s alleged same-sex harassment
involving his supervisor’s request for oral sex).
336. See also Alan David Freeman, Legitimizing Racial Discrimination Through
Antidiscrimination Law: A Critical Review of Supreme Court Doctrine, 62 MINN. L. REV. 1049,
1052–57 (1978) (proposing a “victim perspective” as a replacement of the perpetrator centered
model). See generally Krieger, supra note 96 (describing the prevailing discrimination paradigm
as one premised upon notions of evil intent, and arguing instead for an understanding of much
discrimination as resulting from a morally neutral cognitive bias); Alan Freeman,
Antidiscrimination Law: The View from 1989, 64 TUL. L. REV. 1407, 1411–12 (1990)
(describing, as the “dominant” view in American antidiscrimination jurisprudence, “the
‘perpetrator’ perspective,” which is concerned with “eradicating the behaviors of” those who
have acted with “individual badness”); Charles R. Lawrence III, The Id, the Ego, and Equal
Protection: Reckoning with Unconscious Racism, 39 STAN. L. REV. 317 (1987) (arguing that
unconscious discrimination is widespread and should be recognized as actionable).
426
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
the plaintiff under Title VII337 for maintaining or failing to correct an
abusive working environment that is “severe or pervasive,” the actual
behavior upon which liability is premised is nearly always that of coworkers
or supervisors. Indeed, the “employer” is most often a corporate entity
rather than a physical person. Thus, any inquiry into the discriminatory
motive or intent behind the actual harassment is necessarily one step
removed from the defendant in the case.338 Second, one or a few isolated
instances of harassment generally do not rise to the level of actionable
hostile work environment abuse.339 Instead, in order to be considered
“severe or pervasive” and therefore discriminatory, conduct or abusive
language must usually span a significant period of time and be comprised of
numerous incidents.340 Often, the abusive environment is the result of the
actions of several different persons in the workplace.341 Under these
337. See 42 U.S.C. § 2000e(b) (2003) (stating that “an employer is a ‘person’ with 15 or
more employees”); Newsome v. Admin. Office of the Courts, 51 Fed. Appx. 76 (3d Cir. 2002)
(unpublished opinion); Smith v. Amedisys, Inc., 298 F.3d 434, 448 (5th Cir. 2002) (finding no
individual liability under Title VII); Indest v. Freeman Decorating, Inc., 164 F.3d 258, 262 (5th
Cir. 1999) (holding that Title VII extension of liability to agents of employer was intended to
incorporate the theory of respondeat superior, not to impose liability on individuals).
338. Cf. White & Krieger, supra note 265, at 1195 (making a parallel argument in the
context of challenged employment actions).
339. E.g., Faragher v. City of Boca Raton, 524 U.S. 775, 788 (1998) (stating that “offhand
comments, and isolated incidents (unless extremely serious) will not amount to discriminatory
changes in the ‘terms and conditions of employment’”); Hilt-Dyson v. City of Chicago, 282
F.3d 456, 463 (7th Cir. 2002) (holding that isolated incidents may not constitute sexual
harassment); Brennan v. Metro. Opera Ass’n, 192 F.3d 310, 318 (2d Cir. 1999) (holding that
“[i]solated, minor acts or occasional episodes do not warrant relief”); King v. Bd. of Regents,
898 F.2d 533, 537 (7th Cir. 1990) (explaining that “generally, repeated incidents create a
stronger claim of hostile environment” than single incidents); Breda v. Wolf Camera, Inc., 148
F. Supp. 2d 1371, 1381 (S.D. Ga. 2001) (noting that “merely inserting . . . [a] rude or sexualized
comment/gesture/joke into a lengthy list [of discussions] accumulated over years of
employment does not . . . a Title VII claim make”); Jones v. Clinton, 990 F. Supp. 657, 675–76
(E.D. Ark. 1998) (finding a single, albeit notorious, incident of harassment insufficiently severe
to constitute a hostile environment). But see Tomka v. Seiler Corp., 66 F.3d 1295, 1305 (2d Cir.
1995) (stating that “even a single incident of sexual assault sufficiently alters the conditions of
the victim’s employment and clearly creates an abusive work environment for purposes of Title
VII liability”).
340. In Meritor and subsequent cases, the Supreme Court articulated the standard using the
disjunctive “or.” 477 U.S. at 67. However, courts applying the “severe or pervasive” standard
have generally found that a single or small number of incidents do not satisfy the requirement,
unless that incident is especially heinous (for example rape or serious sexual assault). Id.
341. Indeed, it is precisely this aspect of the law that is troubling to those scholars who
argue that current Title VII hostile environment doctrine raises First Amendment concerns. See,
e.g., Kingsley R. Browne, Title VII as Censorship: Hostile-Environment Harassment and the
First Amendment, 52 OHIO ST. L.J. 481 (1991); Kingsley R. Browne, Zero Tolerance for the
First Amendment: Title VII’s Regulation of Employee Speech, 27 OHIO N.U. L. REV. 563
(2001); Eugene Volokh, Freedom of Speech and Appellate Review in Workplace Harassment
37:0321]
SEXUAL HARASSMENT LAW
427
circumstances, the issue of the intent of the various actors whose behavior
has contributed to the hostile environment becomes problematic.
Individual harassers are not liable for sexual harassment under Title VII.
Though some courts in the past questioned this proposition,342 the
overwhelming weight of authority now holds that it is the employer only
that is liable under the statute.343 The question becomes, therefore, whether
the employer is liable vicariously for the discriminatory intent of the
harassing employees, or rather whether the employer is liable for its own
discriminatory intent. If the latter, then the issue of the motivation of the
harasser or harassers takes on a different and less critical role.
In the hostile environment context, the test for employer liability differs
depending upon whether the harassment is carried out by a supervisor or by
coemployees. Where a supervisor sexually harasses a subordinate, the
employer is liable for the acts of its supervisory employees according to
agency principles of vicarious liability.344 As outlined by the Supreme Court
in Burlington Industries, Inc. v. Ellerth,345 agency principles require that an
employer be strictly liable for “an actionable hostile environment created by
a supervisor with immediate (or successively higher) authority over the
Cases, 90 NW. U. L. REV. 1009 (1996); Volokh, supra note 270, at 627. Professor Volokh points
out that offensive words which in themselves do not constitute a violation of Title VII can,
when considered in combination with other isolated offenses, become actionable. Thus,
employers have a strong incentive to censor all employee speech that might possibly count as a
piece of a larger hostile environment puzzle. Id. at 638–40.
342. See Paroline v. Unisys Corp., 879 F.2d 100, 104 (4th Cir. 1989), rev’d in part and
aff’d in relevant part, 900 F.2d 27 (4th Cir. 1990) (en banc); Goodstein v. Bombardier Capital,
Inc., 889 F. Supp. 760, 763–65 (D. Vt. 1995).
343. See Wathen v. Gen. Elec. Co., 115 F.3d 400, 405 (6th Cir. 1997); Tomka v. Seiler
Corp., 66 F.3d 1295, 1316 (2d Cir. 1995); Gary v. Long, 59 F.3d 1391, 1399 (D.C. Cir. 1995);
EEOC v. AIC Sec. Investigations, Ltd., 55 F.3d 1276, 1282 (7th Cir. 1995); Lenhardt v. Basic
Inst. of Tech., Inc., 55 F.3d 377, 381 (8th Cir. 1995); Smith v. Lomax, 45 F.3d 402, 403 (11th
Cir. 1995); Sauers v. Salt Lake County, 1 F.3d 1122, 1125 (10th Cir. 1993); Grant v. Lone Star
Co., 21 F.3d 649, 651–53 (5th Cir. 1993), cert. denied, 513 U.S. 1015 (1994); Miller v.
Maxwell’s Int’l, Inc., 991 F.2d 583, 587 (9th Cir. 1993), cert. denied, 510 U.S. 1109 (1994);
Busby v. City of Orlando, 931 F.2d 764, 772 (11th Cir. 1991). For a recent article which
criticizes this trend and argues that individual supervisors should be liable for harassment, see
Joanna L. Grossman, The Culture of Compliance: The Final Triumph of Form over Substance
in Sexual Harassment Law, 26 HARV. WOMEN’S L.J. 3, 9, 74 (2003).
344. Courts have relied upon agency principles to justify the imposition of vicarious
employer liability in hostile work environment cases: employers provide supervisory employees
with the authority to make decisions affecting the terms and conditions of employment, and the
employer is therefore responsible for the supervisor’s harassing behavior when it alters those
terms and conditions. See, e.g., Meritor, 477 U.S. at 72 (citing RESTATEMENT (SECOND) OF
AGENCY §§ 219–37 (1958)); Faragher v. City of Boca Raton, 524 U.S. 775, 790–91 (1998);
Burlington Indus., Inc. v. Ellerth, 524 U.S. 742, 759–60 (1998).
345. 524 U.S. 742 (1998).
428
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
employee.”346 This employer liability is absolute in cases in which a
“tangible employment action”347 has been taken, but it is subject to an
affirmative defense in other cases. To prevail on the affirmative defense, the
employer must demonstrate: “(a) that the employer exercised reasonable
care to prevent and correct promptly any sexually harassing behavior, and
(b) that the plaintiff employee unreasonably failed to take advantage of any
preventative or corrective opportunities provided by the employer or to
avoid harm otherwise.”348
Compare this defense to the standard of employer liability in cases of
coworker (nonsupervisor) harassment: In those cases, an employer is not
vicariously liable for the acts of its employees but only for its own actions
according to a negligence standard. Thus, “[a]n employer is negligent with
respect to sexual harassment if it knew or should have known about the
conduct and failed to stop it.”349 It is apparent that, in hostile environment
harassment cases (in which, by definition, no tangible employment action
has been taken), there is no substantive difference in the standard of liability
by which the employer defendant is judged. The questions whether the
346. Id. at 765.
347. In the 2004 term, the Supreme Court resolved an issue that had divided the federal
circuits: whether a constructive discharge is a “tangible employment action” such that it
prevents the employer from relying on the affirmative defense. Penn. State Police v. Suders, 124
S. Ct. 2342, 2550 (2004); see also Robinson v. Sappington, 351 F.3d 317, 336–37 (7th Cir.
2003) (holding that a constructive discharge qualifies as a tangible employment action only
where the action was effected by the official act of a supervisory employee); Reed v. MBNA
Mktg. Sys., Inc., 333 F.3d 27, 33 (1st Cir. 2003) (same); Suders v. Easton, 325 F.3d 432 (3d
Cir. 2003), cert. granted, 504 U.S. 1040 (2003); Jaros v. LodgeNet Entm’t Corp., 294 F.3d 960
(8th Cir. 2002) (holding that a constructive discharge constitutes a tangible employment action
and thus bars the affirmative defense); Turner v. Dowbrands, Inc., No. 99-3984, 2000 WL
924599, at *1 (6th Cir. June 26, 2000) (unpublished decision) (holding that a constructive
discharge does not constitute a tangible employment action and thus does not prevent the
defendant from asserting the affirmative defense); Caridad v. Metro-North Commuter R.R., 191
F.3d 283, 294–95 (2d Cir. 1999), cert. denied, 529 U.S. 1197 (2000) (same). In Suders, the
Supreme Court held that a constructive discharge is not a tangible employment action for
purposes of the availability of the Ellerth/Faragher affirmative defense except in cases in which
that action is effectuated by a supervisor’s official act. 124 S. Ct. at 2344, 2355–56.
348. Ellerth, 524 U.S. at 765. Though this defense is framed in the conjunctive, it has
sometimes been applied in the alternative. See Grossman, supra note 341, at 8 n.23 (discussing
cases in which courts have eliminated the second prong of the defense).
349. Watson v. Blue Circle, Inc., 324 F.3d 1252, 1257 (11th Cir. 2003) (to face liability,
employer must have had notice of the harassment and failed to remedy the problem); Ellerth,
524 U.S. at 759; see also Courtney v. Landair Transp., Inc., 227 F.3d 559, 564 (6th Cir. 2000)
(the master-servant relationship does not apply in coworker cases; rather a negligence standard
applies); 29 C.F.R. § 1604.11(d) (2003) (“With respect to conduct between fellow employees,
an employer is responsible for acts of sexual harassment in the workplace where the employer
(or its agents or supervisory employees) knows or should have known of the conduct, unless it
can show that it took immediate and appropriate corrective action.”).
37:0321]
SEXUAL HARASSMENT LAW
429
employer “exercised reasonable care to prevent and correct promptly” the
harassing behavior on the one hand, or “knew or should have known about
the conduct and failed to stop it” on the other, are essentially the same. The
practical difference, then, lies in which party bears the burden of proof in
demonstrating the employer’s reasonableness or lack thereof.
The substantive issue that underlies the question whether an employer is
liable for a hostile environment created by its employees, whether
supervisory or not, is the reasonableness of the employer’s conduct in the
particular case. Under these circumstances, it makes little sense to inquire
very far, if at all, into the subjective intent of the individual harasser or
harassers. The subjective intent of the harasser should play a lesser role
because it is the reasonableness of the employer’s conduct that is at issue.
Even where the employer acts through its employees, its liability depends
on its actions taken in response to the creation of the hostile environment
(or to prevent it), and not on its actions as taken by the harassing
employees.350 Under these circumstances, it is both reasonable and fair to
charge employers with the knowledge351 that the patterns of behaviors
generally involved in hostile environment harassment cases, as outlined in
Part IV, occur “because of . . . sex” as required by Title VII.
A focus on the reasonableness of the employer’s response to hostile
environment harassment also makes much sense in those cases in which the
hostile environment is created by the combined effect of the conduct of
multiple employees. It would be difficult, and would entail engaging in a
legal fiction in any event, to attempt to discern some “collective” intent for
a hostile environment created in this way.352 As courts have repeatedly
350. The exception to this analysis is the “pure” quid pro quo case in which a supervisor
takes a tangible job action based upon the employee’s acceptance or rejection of a sexual quid
pro quo. There one might argue that the intention of the individual supervisor is itself quite
relevant. It is precisely these cases, however, in which evolution theory most clearly
demonstrates that the harasser’s behavior is driven by the sex of the victim.
351. In his opinion in Faragher v. City of Boca Raton, 524 U.S. 775 (1998), decided on the
same day as Ellerth, Justice Souter engaged in a remarkably candid discussion of the reasons for
vicarious liability of employers for harmful acts of their agents. The opinion states, “In the
instances in which there is a genuine question about the employer’s responsibility for harmful
conduct he did not in fact authorize, a holding that the conduct falls within the scope of
employment ultimately expresses a conclusion not of fact but of law.” Id. at 796. Thus, in
determining the scope of employer liability for harassing conduct, the court should evaluate “the
reasons that would support a conclusion that harassing behavior ought to be held within the
scope of a supervisor’s employment, and the reasons for the opposite view.” Id. at 797.
352. Several commentators have considered the nature of the “because of” element of Title
VII. The issue is generally framed as an inquiry into whether Title VII requires a conscious
motivation to treat members of a protected class differently, or rather whether some degree of
causation-in-fact is all that is required. See Robert W. Belton, Causation and Burden-Shifting
Doctrines in Employment Discrimination Law Revisited: Some Thoughts on Hopkins and Wards
430
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
emphasized, a hostile working environment may arise out of an amalgam of
individual words and actions, any one of which might be unfortunate or
obnoxious or offensive, but not itself actionable under Title VII.353 Under
these circumstances, it is much more useful to answer the question of
discriminatory intent by virtue of the conduct of the defendant-employer.
Cove, 64 TUL. L. REV. 1359, 1383–84 (1990) (discussing the differential aims of judicial
inquiries into causation and intent in Title VII hostile work environment cases); Paul J. Gudel,
Beyond Causation: The Interpretation of Action and the Mixed Motives Problem in Employment
Discrimination Law, 70 TEX. L. REV. 17, 52 (1991) (suggesting that courts have found the
causation model to be an “unsatisfactory formulation” of the plaintiff’s burden of proof, and that
courts have consequently “been left free to fall back on a purely intuitive sense of what evidence
establishes a sufficient causal nexus between motive and act”); George Rutherglen,
Reconsidering Burdens of Proof: Ideology, Evidence, and Intent in Individual Claims of
Employment Discrimination, 1 VA. J. SOC. POL’Y & L. 43, 48–49 (1993) (explaining that Title
VII’s “because of” language actually refers to an employer’s course of action “based on” a
prohibited characteristic, and that courts and commentators have erroneously conflated this
requirement with an inquiry into the employer’s subjective intent). Professor Linda Krieger and
others have argued persuasively that differential treatment of racial or ethnic minorities or of
women can be based upon unconscious or unintentional cognitive processes and that factual or
but-for causation is therefore a more appropriate inquiry under Title VII. Krieger, supra note
96; Lawrence, supra note 336; McGinley, supra note 265. As to the nature of the causation
standard, many scholars advocate a “but/for” approach. See id. at 419 (insisting that Title VII’s
“proof mechanisms serve the role of determining causation rather than conscious intent,
assuring that the underlying employment decision is made because of the employee’s protected
characteristic, either with or without the employer’s conscious awareness”) (emphasis omitted);
Jeffry A. Van Detta, “Le Roi Est Mort; Vive Le Roi!”: An Essay on the Quiet Demise of
McDonnell Douglas and the Transformation of Every Title VII Case After Desert Palace, Inc. v.
Costa Into a “Mixed-Motives” Case, 52 DRAKE L. REV. 71, 93–94 (2003) (arguing that Title
VII’s “because of” language mandates a consideration of “but/for” causation rather than
“intent,” because the requisite intent is “always outlying, lurking in the background in our
discriminating society”). The statutory and judicial approaches to the “mixed-motives” scenario
tend to support the argument in favor of a but-for approach to causation, at least in cases
involving tangible employment decisions. In a mixed motives case, the plaintiff claims that a
discriminatory purpose was part of the motivation for a particular employment decision, though
other legitimate motives were also present. The 1993 amendment to Title VII codified the
mixed motive claim and provided for an affirmative defense. 42 U.S.C. § 2000e-2(m) (2000). If
the plaintiff demonstrates that a discriminatory motive entered into the employment decision,
the employer then has an opportunity to defend against the claim by demonstrating that the
decision would have been the same even absent the improper motive. 42 U.S.C. § 2000e5(g)(2)(B)(ii). The Supreme Court recently held that the plaintiff can satisfy its burden of
demonstrating that an unlawful motive entered into the decision-making process wholly with
circumstantial evidence. Desert Palace, Inc. v. Costa, 539 U.S. 90, 99–100 (2003).
353. E.g., Faragher, 524 U.S. at 788; Tomka v. Seiler Corp., 66 F.3d 1295, 1305 (2d Cir.
1995) (holding that verbal harassment standing alone is not enough to create an abusive
working environment; rather, incidents of harassment must occur in concert or with a regularity
that can be reasonably termed pervasive); Snell v. Suffolk County, 782 F.2d 1094, 1103 (2d Cir.
1986) (requiring evidence of more than just “a few isolated incidents of racial enmity”). See
generally Volokh, supra note 270, at 647 (arguing that isolated instances of harassment by
many persons may, in the aggregate, be severe and pervasive).
37:0321]
SEXUAL HARASSMENT LAW
431
In Davis v. Monroe County Board of Education,354 the Supreme Court
confronted an analytically identical issue.355 There, a student alleged that
she had been sexually harassed by another student, and she attempted to
hold the school board liable for its failure to remedy the situation in the face
of repeated complaints and requests.356 The Court stated: “We disagree with
respondents’ assertion . . . that petitioner seeks to hold the Board liable for
[the harasser’s] actions instead of its own. Here, petitioner attempts to hold
the Board liable for its own decision to remain idle in the face of known
student-on-student harassment in its schools.”357 Without examining the
specific intent of the child whose behavior was alleged to have created the
abusive environment, the Court assumed that the harassment, if severe and
pervasive so as to compromise the victim’s equal access to an educational
program or activity, would be actionable and that the school board would be
liable for its deliberate indifference to the harassment.358
Similarly, in sex discrimination cases that fall within one of the behavior
patterns outlined in Part III, which based on evolutionary analysis generally
occur because of both the sex of the harasser and the sex of the victim, the
causation element of section 703 should be presumed satisfied insofar as the
intent of the individual harasser is concerned. This is because, on a systemic
level, these types of behavior occur because of sex and the employer should
be held to this knowledge when the reasonableness of its own behavior is
evaluated.359 Insofar as both Congress and the Supreme Court have made
354. 526 U.S. 629 (1999).
355. In Davis, the claim was brought under Title IX rather than Title VII. Id. at 632–33.
Thus, the standard by which the defendant was potentially liable for its failure to prevent or
remedy the severe student-on-student harassment alleged was different from the negligence
standard applicable under Title VII. Under Title IX, the appropriate inquiry is whether the
funding recipient has acted with “deliberate indifference.” See Bryant v. Indep. Sch. Dist., 334
F.3d 928, 934 (10th Cir. 2003) (setting a standard to be deliberate indifference for harassment);
C.R.K. v. U.S.D. 260, 176 F. Supp. 2d 1145, 1163 (D. Kan. 2001) (finding school not
deliberately indifferent, therefore not liable for harassment). The majority in Davis explicitly
drew on Title VII sexual harassment cases in its discussion. 526 U.S. at 651 (citing Oncale, 523
U.S. at 75 and Meritor, 477 U.S. at 57).
356. Davis, 526 U.S. at 641.
357. Id.
358. See id. at 648–53.
359. A burden-shifting scheme similar to that employed in the pretext cases might usefully
be employed in hostile environment cases in which the employer wishes to assert that, despite
the general pattern, the conduct in the particular case was not based on the plaintiff’s sex. Cf.
McDonnell Douglas Corp. v. Green, 411 U.S. 792, 802–03 (1973). This approach, consistent
with the vicarious liability formulation of Ellerth and Faragher, would place upon the
defendant the burden of demonstrating that harassing behavior was not based on the sex of the
plaintiff, once the plaintiff had offered prima facie evidence that that the behavior was based
upon her sex. This prima facie case would be satisfied by evidence that the conduct fell within
432
ARIZONA STATE LAW JOURNAL
[Ariz. St. L.J.
clear that prevention and deterrence are of primary importance in the Title
VII statutory scheme,360 such an approach would shift the focus of the
analysis from the spandrel of individual blameworthiness to the functional
doctrinal element of sex-based discrimination and harm.
VI. CONCLUSION
Evolutionary theory has the potential to aid understanding of various
legal issues both from a metaphorical standpoint and as a tool for
understanding broad patterns of human behavior. As metaphor, the concept
of evolutionary spandrels highlights the tendency of functionally more
marginal doctrinal elements to assume heightened importance in the minds
of judges and scholars. In this way, the spandrels come to be viewed as
central to the doctrinal structure when in fact they should more properly be
seen as peripheral. The complex development of Title VII sexual
harassment law has resulted in several of these doctrinal spandrels: a focus
on the sexual character of the harassing behavior rather than on the
functional motives of the harasser(s); an attempt to hold racial and sexual
harassment doctrine consistent over the broad sweep of discrimination
cases; and an undue focus on the intent and blameworthiness of individual
harassers rather than on the employer’s actions in preventing and addressing
reasonably foreseeable harassment based on sex. Both doctrine and theory
would be well-served by an effort to refocus the inquiry upon the more
central, functional, and weight-bearing elements of sexual harassment
doctrine.
In the arena of sex discrimination law, evolutionary analysis as a
substantive means of understanding patterns of human behavior has been
utilized primarily by scholars who argue that Title VII is misguided.
However, evolutionary understandings of the behaviors often seen in
harassment cases suggest that these patterns of behavior are based upon the
sex of the victim and thus satisfy the “because of . . . sex” requirement of
Title VII. Insofar as evolutionary analysis of gender difference has been
employed by those who would criticize laws that prohibit sexual harassment
in the workplace, it is entirely appropriate to ask whether such evolutionary
theory might in fact support a feminist conception of sexual harassment. An
one of the patterns understood by evolutionary theory to be generally based upon the sex of the
victim.
360. See, e.g., Kolstad v. Am. Dental Ass’n, 527 U.S. 526, 545 (1999) (stating that “[Title
VII’s] ‘primary objective’ is ‘a prophylactic one,’” and that “it aims, chiefly, ‘not to provide
redress but to avoid harm’”) (quoting Faragher v. City of Boca Raton, 524 U.S. 775, 806
(1998); Albemarle Paper Co. v. Moody, 422 U.S. 405, 417 (1975)).
37:0321]
SEXUAL HARASSMENT LAW
433
evolutionary perspective on harassing behaviors in the workplace suggests
that the broad patterns seen in typical harassment cases occur, on a systemic
level, because of the victim’s sex. Thus, where such patterns are present,
employers should be presumed to know that such conduct is discriminatory
on the basis of sex. A standard of reasonableness would therefore require an
employer, the only party liable under Title VII, to take steps to prevent and
correct those environments likely to bring about the harassing behaviors
seen in the typical classes of cases analyzed herein.
This approach neither excuses harassment by individuals nor suggests
that sexual harassment is inevitable because the behavior is somehow
genetically determined. Rather, it recognizes that certain work environments
are more likely, on average, to give rise to socially harmful and
discriminatory patterns of behavior. Furthermore, it illuminates the manner
in which sexual harassment tends to constrain female autonomy and to
restrict access to resources in order to constrict female agency and choice.
This recognition, in turn, puts the onus on employers to be proactive in
deterring and preventing such presumptively discriminatory behaviors.
Scientific study of male/female difference will no doubt continue whether
legal feminists embrace or reject it. But if evolutionary analysis can help to
accomplish the goals of reducing workplace sexual harassment and
promoting female autonomy, surely it merits serious discussion. At the very
least, it is incumbent upon those scholars who have employed evolutionary
analysis for the purpose of criticizing sexual harassment law to consider that
evolutionary theory might in fact illuminate the discriminatory nature of
typical workplace harassment behaviors. Even better, perhaps engagement
across these issues will create a dialogue that could bridge gaps between
feminists and evolutionary psychologists and carry the debate forward as
science continues to add to our understandings of human behavior.